- Email: [email protected]
Pebbly versus bouldery rock glaciers: Morphology, structure and processes
Geomorphology 73 (2006) 279 – 296 www.elsevier.com/locate/geomorph
Pebbly versus bouldery rock glaciers: Morphology, structure and processes Atsushi Ikeda a,b,*, Norikazu Matsuoka c a
b
Division of Earth Sciences, National Institute of Polar Research, Itabashi, Tokyo 173-8515, Japan Water and Environmental Research Center, Institute of Northern Engineering, University of Alaska-Fairbanks, Fairbanks, Alaska 99775-5860, USA c Graduate School of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8572, Japan Received 20 January 2005; received in revised form 21 July 2005; accepted 21 July 2005 Available online 26 September 2005
Abstract Differences in rockwall geology cause two types of rock glaciers: a bouldery rock glacier, having the active layer composed of matrix-free boulders derived from crystalline rocks and massive limestone; and a pebbly rock glacier, consisting of matrix-supported debris derived from less resistant shale and platy limestone. Such material composition controls transport processes responsible for the shape, size and internal structure of the two types of rock glaciers. This paper compares the major processes controlling the morphology and structure of the two types in the Swiss Alps, based on mapping, description of geological and morphological characteristics, direct observation of stratigraphy and geophysical soundings of internal structure. In the Swiss Alps, pebbly rock glaciers are distinguished from bouldery rock glaciers by the clast size of 15–20 cm in the mean b-axis diameter. The former are fed by small-scale rockfalls, debris flows and solifluction, whereas the latter mainly originate from large rockfalls. Pebbly rock glaciers are generally smaller (b 200 m in length) than bouldery rock glaciers, because the small exposure of the less resistant source rockwall (b50 m in height) strongly constrains debris supply. As a result, pebbly rock glaciers usually terminate within a valley-side slope, whereas many bouldery rock glaciers extend into the valley bottom. The location of pebbly rock glaciers results in the low frontal slope (V 20 m) and the lack of transverse ridge-furrow topography, because of the lack of compressive flow. The pit-borehole stratigraphy and low DC resistivities (b 10 kVm) indicate ice-cemented or slightly supersaturated permafrost in the pebbly rock glaciers, which presumably originates from groundwater freezing. In contrast, highly ice-supersaturated structure indicated by high DC resistivities (N 100 kVm) in bouldery rock glaciers appears to originate from snow banks buried with deposits of large rockfalls from the large source rockwall. D 2005 Elsevier B.V. All rights reserved. Keywords: Rock glacier; Mountain permafrost; Debris supply; Geophysical soundings; Rock control; Swiss Alps
1. Introduction
* Corresponding author. Division of Earth Sciences, National Institute of Polar Research, Itabashi, Tokyo 173-8515, Japan. Tel.: +81 3 3962 4695; fax: +81 3 3962 5741. E-mail address: [email protected] (A. Ikeda). 0169-555X/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.geomorph.2005.07.015
The dynamics of rock glaciers reflects geology of the source rockwall, which controls the manner of debris supply. A few studies have highlighted the geological constraint on the development of rock glaciers (Wahrhaftig and Cox, 1959; Evin, 1987; Chueca,
280
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Fig. 1. Typical bouldery rock glaciers in the study area. (a) Murte`l I (MT) rock glacier having well-developed ridges and furrows and rimmed with a steep frontal slope 20 m in height. The surface boulders consist of gneiss and schist. (b) PN4 and PN5 rock glaciers originating from a massive limestone rockwall. The total height of the front reaches about 50 m.
1992). These studies state that rock glaciers mainly develop in lithology producing blocky debris, whereas they are infrequent on slopes composed of platy debris with finer materials. The size of clasts composing rock glaciers varies significantly with lithology. Evin (1987) has found that most rock glaciers in the southwestern Alps consist of coarse blocky debris derived from resistant rockwalls (granite, gneiss, sandstone and massive limestone), whereas rock glaciers derived from less resistant schistose rockwalls contain a large
amount of fine debris and small pore volume. Other studies have examined the volumetric relationships between rockwalls and rock glaciers (e.g. Barsch, 1977; Frich and Brandt, 1985; Humlum, 2000). Details are rarely understood, however, on the processes that the source rockwall contributes to the dynamics of rock glaciers. Here, we focus on the material composition of rock glaciers, which reflects the lithology of the source rockwall and controls slope processes. In this respect, we
Fig. 2. Pebbly rock glaciers in the study area. (a) BN rock glacier consists of the upper lobe (BNU) and lower lobe (BNL). The surface clasts are mainly shale pebbles and cobbles. BNL lacks a steep frontal slope. (b) NN10-12 rock glaciers, consisting mainly of shale pebbles and cobbles. (c) NN2 rock glaciers lying on the middle of a talus slope. (d) Exceptionally long (740 m) pebbly rock glacier A7, flowing along a valley bottom.
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Fig. 3. The location of the study area.
classify rock glaciers in terms of composition of surface materials into two types (Matsuoka et al., 2005). A bouldery rock glacier has an active layer composed of matrix-free boulders (Fig. 1). Most of the previously reported rock glaciers belong to this type. In contrast, a pebbly rock glacier is covered with pebbles and cobbles, mostly supported by sandy and silty matrix between coarse materials (Fig. 2). The latter type is probably comparable to the earthy rock glacier defined by Evin (1987). This paper compares the two types of rock glaciers in the Swiss Alps in terms of morphology and structure.
281
Field investigations involved mapping, description of geological and morphological characteristics, excavation with portable drills, seismic refraction soundings and one- and two-dimensional electrical resistivity soundings. These investigations mainly featured pebbly rock glaciers that have been poorly described, whereas morphology and structure of bouldery rock glaciers in the study area have already been well documented (e.g. Ka¨a¨b et al., 1998; Frauenfelder and Ka¨a¨b, 2000; Ikeda and Matsuoka, 2002). Ice is another vital component of rock glaciers, although its origin is problematic. In the case where a large rock glacier develops from a cirque floor or the head of a U-shaped valley, glacial, periglacial and their continuum origins have been proposed (e.g. Outcalt and Benedict, 1965; Giradino and Vitek, 1988; Humlum, 2000). In this paper, subsurface perennial ice in rock glaciers is called permafrost regardless of the origin because we focus on talus-derived rock glaciers that have had no contact with glaciers during the Holocene (cf. Frauenfelder et al., 2001). The development of nonglacial ice in rock glaciers is also discussed by comparing the structure and processes of pebbly and bouldery rock glaciers.
Fig. 4. Contour map (a) and geological map (b) of the study area, shown with the distribution of rock glaciers.
282
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
2. Study area Field investigations were undertaken in the Upper Engadin in the Swiss Alps, where a large extent of high mountain slopes permits the widespread occurrence of permafrost (Figs. 3 and 4a). The timberline lies at 2000–2200 m ASL, the mean annual air temperature 08C isotherm at 2200–2300 m ASL and the glacial equilibrium line at about 3000 m ASL (Haeberli et al., 1992). In contrast with the limited extent of glaciers, rock glaciers dominate mountain areas with peaks around 3000 m ASL (e.g. Frauenfelder et al., 2001; Ikeda and Matsuoka, 2002). About 80% of the active rock glaciers are considered to be talus-derived (Frauenfelder et al., 2003). Glacial erosion during the Pleistocene basically determined the morphology of mountain slopes in the study area. Bedrock lithology appears to primarily con-
trol the resistance of the slopes to glacial and paraglacial erosion. The bedrock in this region is broadly divided into the crystalline rock group (granite, granodiorite, diorite, gneiss and schist) and the sedimentary rock group (limestone, dolomite, conglomerate and shale) (Fig. 4b). The crystalline rocks generally sustain higher rockwalls than the sedimentary rocks (Matsuoka and Ikeda, 2001). The bedrock lithology also affects the size of debris: massive limestone, conglomerate and most of the crystalline rocks produce coarse debris, whereas platy/porous limestone and shale favour production of fine debris. The difference in size of the debris affects periglacial landforms. In general, coarse debris constructs talus slopes and rock glaciers below high rockwalls, whereas fine debris promotes development of solifluction lobes and pattern ground on gentle slopes (Matsuoka et al., 1997) and, where debris sedimentation is sufficient, can produce rock glaciers.
Fig. 5. Distribution of rockwalls and rock glaciers in the Corviglia region. (a) Northern area. (b) Southern area. Small-scale topography and the sites of geophysical soundings are also indicated.
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
The major study area is Corviglia, a large ski region north of St. Moritz (Fig. 5). In the northern Corviglia area, granite and limestone are the major lithologies composing rockwalls and shale is partly exposed on the flank of Piz Padella and Las Trais Fluors (Fig. 5a). In the southern Corviglia area, conglomerate composes the southern face of Piz Nair, whereas shale partly embedding chert composes the northern face (Fig. 5b). The high shale rockwall of Piz Nair is supported by the overlying conglomerate caprock. The morphology of pebbly rock glaciers were also studied on the northern face of Piz Blaisun, lying north of Albulapass, and the northern area of Julierpass (Fig. 4). The former area is mainly composed of platy limestone, and the latter massive limestone. Both areas also include shale outcrops. 3. Methodology 3.1. Distribution and morphology Bouldery and pebbly rock glaciers were mapped on the 1/25,000 topographic maps of Switzerland on the basis of field and air photograph surveys. Previous information on the distribution of rock glaciers in the Upper Engadin by Hoelzle (1998) and Frauenfelder et al. (2001) were also incorporated in the mapping. The lithologies of these rock glaciers were determined in the field with the help of the 1/200,000 geological map of Engadin. Small-scale topography, such as ridges and furrows, lobate bulges and debris flow levees on and around rock glaciers, was also mapped on the 1/15,000 farmland-name map of Celerina, which covers the Corviglia area. Morphological parameters of 28 pebbly rock glaciers were measured on the maps and in the field. The parameters include the aspect, frontal altitude, length, width, frontal height, surface gradient and maximum height of headwall. The frontal height was defined as the height of the steep frontal slope distinguished from the upper surface by a nick-point. In the field, longitudinal profiles of two bouldery and 18 pebbly rock glaciers were determined with a clinometer and an ultrasonic distance meter. 3.2. Surface composition The prevailing clast size was evaluated on 25 pebbly rock glaciers. In addition, the size of the surface clasts was more accurately determined on ten rock glaciers and talus slopes along the longitudinal profile. The measurement sites included four pebbly rock glaciers
283
consisting of shale (BNU, BNL, BW2, NN8) and four bouldery rock glaciers consisting of granite (O1, O2) or massive limestone (PN5, PS3) in the Corviglia area (see Figs. 4a and 5 for the locations). The measurement was also carried out on two rock glaciers consisting of gneiss and schist near the Corviglia area: Murte`l (MT) rock glacier lying 8 km south of St. Moritz and Muragl (MR) rock glacier lying 6 km east of St. Moritz, where intensive observations have been undertaken (e.g. Haeberli et al., 1999; Arenson et al., 2002). On each profile, four to eight horizontal survey lines were chosen. These lines were located longitudinally at 18– 130 m intervals. For each line, 25 clasts were sampled at intervals of 0.5, 1 or 2 m and the three-axis diameters were measured. The interval depended on the predominant size of the clasts. The size of rockfall debris is controlled by the joint spacing of the source rockwall (Matsuoka and Sakai, 1999). Average joint spacing was also measured at the foot of rockwalls above the talus slope. The number of joints, crossing a ring wire 2 m in circumference, was counted on the rock face and divided by 2 m to compute the average joint spacing. The next measurement was made by shifting the wire by 0.6–0.7 m, and in total 10 measurements were performed at one site. Three sites were chosen on a rockwall. As a result, the average joint spacing on the rockwall represented the mean value for 30 points. 3.3. Internal structure 3.3.1. Direct observation The internal structure of the BNU pebbly rock glacier was directly observed by excavation and drilling in early August 2000. A 3-m deep pit was excavated with a shovel and rock-cutting machine and then an additional 2.4 m deep borehole (6 cm in diameter) was drilled with an engine auger down from the bottom of the pit. The stratigraphy was observed on the pit wall and with the borecores. The gravimetric ice content was determined for the borecores at 4 and 5 m depths. 3.3.2. Seismic refraction soundings The presence of permafrost was investigated in eight pebbly and three bouldery rock glaciers by seismic refraction soundings. A high velocity layer (z 2000 m s 1) is regarded to be frozen where it underlies a low velocity layer (b 1500 m s 1) much thinner than the apparent thickness of the rock glacier (e.g. Hunter, 1973; Haeberli and Patzelt, 1982). A 3-channel seismograph, McSEIS-3 (manufactured by Oyo, Japan) was used for the survey. A 4-kg sledge-
284
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
permafrost and/or between layers having different grain size distribution, DC resistivities reinforce the interpretation of the internal structure of rock glaciers suggested by borecores or other geophysical properties (e.g. Fisch et al., 1977; Haeberli and Vonder Mu¨hll, 1996; Vonder Mu¨hll et al., 2000). In particular, the resistivity value is an indicator of unfrozen water content in permafrost, which mainly reflects ground temperature and grain-size distribution. One-dimensional (vertical) electrical DC resistivity soundings were performed on two bouldery and eleven pebbly rock glaciers from late July to early August in 1999 and 2002 with the resistivity meter SYSCAL R1 PLUS (manufactured by Iris Instruments, France). The setting of the electrodes followed the Schlumberger array. The resistivity values of the subsurface layers and the boundaries between these layers were calculated with the software ElecAuto (Oyo, Japan).
hammer produced a seismic pulse. The shot points were placed at 2 or 2.5 m intervals near the receivers and otherwise at 5 or 6 m intervals. A survey line was 40– 60 m long. The P-wave velocities of two or three layers and the depth of the layer boundaries were determined after a correction with regard to the dip of the boundary (Palmer, 1986). The soundings were carried out from middle July to middle August in 2000–2003. 3.3.3. DC resistivity soundings The structure in rock glaciers was investigated by direct current (DC) resistivity soundings. Resistivity in ground increases when ground temperature falls below 0 8C because of a decrease in liquid water content (e.g. Hoekstra and McNeill, 1973). Thus, if water content is homogeneous in the ground, permafrost is detected by DC resistivity soundings. Although the water content varies between the active layer and Table 1 Morphology and composition of pebbly rock glaciers Site
Vegetation1
A1U A1L A3 A4U A4L A5U A5L A6 A7 A8 BNU BNL BN2 BN3U BN3L BW1 BW2 NN1 NN2 NN3 NN4 NN8 NN9 NN10 NN11 NN12 J1 J2
NV NV NV NV NV NV NV NV NV NV NV NV NV PV DV NV NV DV NV PV PV PV DV NV NV NV NV NV
Slope aspect
Frontal altitude
Length
Width
Frontal slope ht.
Surface gradient
(8)
(m)
(m)
(m)
(m)
(8)
65 65 30 355 355 15 15 290 275 336 35 35 40 340 350 270 270 355 5 5 10 35 70 35 20 30 90 275
2570 2540 2540 2645 2615 2650 2605 2670 2550 2385 2800 2765 2730 2640 2610 2790 2810 2570 2605 2620 2640 2705 2710 2735 2755 2770 2835 2735
45 75 55 35 55 55 85 505 740 500 50 100 45 90 50 110 90 35 90 55 140 240 80 245 290 225 40 90
110 70 100 85 70 70 110 160 140 180 130 105 120 230 165 100 90 100 65 110 130 100 120 100 90 200 180 40
16 4 9 34 9 11 14 20 25 20 10 0 20 4 3 5 13 9 14 8 20 10 5 7 na* na* 10 3
23 21 18 24 25 23 21 15 17 21 25 20 18 28 18 23 28 13 30 21 17 17 19 20 15 18 15 29
Lithology2
Dominant clast size
Location3
(cm) LSp, SH LSp, SH LSp LSp LSp LSp LSp LSp, LSm LSp, LSm LSp, LSm SH, LSm SH, LSm LS m, SH SH, LSm SH, LSm SH, LSm SH SH SH, CG SH, CG CG, SH SH SH SH, CG SH, CG SH, CG LSp SH
5–10 5–10 10–20 10–20 10–20 10–20 10–20 15–25 15–25 5–10 5–15 N20 5–10 5–10 3–8 3–8 3–8 5–10 N20 5–10 5–15 5–15 5–15 10–20 5–10
Headwall max. height
Other features4
(m) TM TF TF TM TM TM TF VF VF VF TM TF TF TM TF TF TF TF TM TF TF TF VF TF TF TF TF TM
b5 – – b5 – 40 – 200 340 320 30y – 40 10y – 30y b5 50 100 50 20 150 – 160 190 160 10 b5
OW OW OW OW OW RF RF RF
OWB RF
RF
OWB RF
OW
1 NV = not or sparsely vegetated (probably active); PV = partly vegetated (probably inactive); DV = densely vegetated (probably relict). 2 LSp = platy limestone/dolomite; LSm = massive limestone/dolomite; SH = shale; CG = conglomerate. 3 TF = foot of a talus slope; TM = middle of a talus slope; VF = valley floor. 4 OW = neither boulder nor sandy–silty matrix near the surface; OWB = a thin openwork bouldery layer on the surface; RF = transverse ridges and furrows on the surface. *Artificially modified front. yThe height of a caprock consisting of massive limestone: the underlying shale rockwall is lower than 5 m.
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
The two-dimensional internal structure of two bouldery and four pebbly rock glaciers was estimated by the DC resistivity tomography method (DC resistivity imaging). The electrode configuration followed the Wenner–Schlumberger array, which combines the Wenner and Schlumberger configurations in a profile (e.g. Ishikawa et al., 2001). The sounding profiles were set parallel to the flow line, on which electrodes were placed at 4 or 5 m intervals. The soundings were carried out with the resistivity meter SYSCAL JUNIOR (Iris Instruments, France) from middle July to early August in 2001 and 2002. Two-dimensional modelled DC resistivity distributions were computed with the software RES2DINV ver. 3.4 produced by M. H. Loke.
285
4. Surface features 4.1. Distribution and lithology In the study area, pebbly rock glaciers occur on slopes dominated by shale and/or platy limestone debris (Table 1, Fig. 4b). In contrast, bouldery rock glaciers are located below rockwalls consisting of crystalline or massive sedimentary rocks. Because the areas of shale and platy limestone are restricted, fewer pebbly rock glaciers exist than bouldery rock glaciers (Fig. 4b). Eighty-five percent of the pebbly rock glaciers are located on valley side slopes, reflecting the small size (Table 1, Figs. 2a–c and 5). In particular, the front of three pebbly rock glaciers (A4L, NN2, J2) has not
Fig. 6. Longitudinal profiles of pebbly rock glaciers. A profile of a small but typical bouldery rock glacier is also shown. The vertical scale is elongated to twice the horizontal scale.
286
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Fig. 7. Relationship between the length of pebbly rock glaciers and the maximum height of the source rockwalls. NV=not or sparsely vegetated (probably active) and PV=partly vegetated (probably inactive).
reached the foot of a talus slope (Fig. 2c). Exceptionally long (N 500 m) pebbly rock glaciers (A6, A7, A8) develop in the valley bottom (Fig. 2d). 4.2. Length of a rock glacier and rockwall height The length of rock glaciers relates to the height of the source rockwalls. Seventy-five percent of the pebbly rock glaciers are shorter than 200 m (Table 1, Figs. 5 and 6). Half of such short rock glaciers compose multiple lobes (A1U, A1L, A4U, A4L, A5U, A5L,
BNU, BNL, BN3U, BN3L), but the total length of each lobe is still shorter than 200 m. The source rockwalls of such short rock glaciers are mostly lower than 50 m (Fig. 7). These dimensions contrast with those of active bouldery rock glaciers in the study area, 70–80% of which are longer than 200 m (Hoelzle, 1998; Matsuoka et al., 2005). The source rockwalls are generally higher than 100 m. Pebbly rock glaciers, longer than 200 m (A6, A7, A8, NN8, NN10, NN11, NN12), develop below rockwalls higher than 150 m, where the conglomerate caprock or inclusion of massive limestone can support the high rockwalls (Table 1, Fig. 7). Below a large source rockwall (340 m high), A6 and A7 rock glaciers, the former partly covering the upper part of the latter, constitute a sequence of exceptionally long (c. 1 km) pebbly rock glaciers along a valley floor. 4.3. Thickness and surface gradient of a rock glacier The observed pebbly rock glaciers generally have a low frontal slope compared with the previously known (bouldery) rock glaciers (cf. Burger et al., 1999). The height of frontal slope of the former is smaller than 25 m, except for A4U which lies on a steep slope (Table 1, Fig. 6). Half of the frontal slopes are lower than 10 m. In contrast, active bouldery rock glaciers in the study area mostly have a frontal slope higher than 20 m (Matsuoka et al., 2005). Even the smallest bouldery rock glaciers (PN4 and PN5) have high frontal slopes (c. 20 m: Fig. 6). These results suggest that the pebbly rock glaciers are composed of thinner sediments than
Fig. 8. Downslope variation in the average clast size from the top of a talus slope to a rock glacier. The symbols indicate the clast size on the rock glacier (a) and the talus slope (b). The symbol at 0 m distance shows the average joint spacing of the source rockwall. The lines connect the values on the same survey profile. The solid line (shale) indicates pebbly rock glaciers and the broken line (other lithologies) bouldery rock glaciers. Data for BNU and BNL were obtained along a survey profile.
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
the bouldery rock glaciers, although the frontal height of the former tends to slightly underestimate the thickness, because the upper surface gradually inclines toward the front. The surface gradient of the pebbly rock glaciers ranges from 138 to 298 (average 218), which mostly overlaps with that of bouldery rock glaciers (Matsuoka et al., 2005). The longitudinal profiles of the upper surface of the pebbly rock glaciers, however, are different from those of typical bouldery rock glaciers. The profile of pebbly rock glaciers is convex-up (A1U, A1L, BNL, BW1, NN2, NN10, NN12) or almost flat, whereas that of a typical bouldery rock glacier (PN5) is concave-up (Fig. 6). 4.4. Surface topography The ridge-furrow topography, which is common on bouldery rock glaciers (e.g. Burger et al., 1999), develops only on six pebbly rock glaciers with surface gradient of less than 208 (A6, A7, A8, BN3L, NN1, NN9) (Table 1, Figs. 2 and 5). Despite the absence of the ridge-furrow topography, the upper surface of some pebbly rock glaciers (NN10, NN11, NN12) is undulated because of lobate bulges (Figs. 5b and 6). These bulges may represent a flow unit moving faster than the lower part. Such bulges also develop on talus slopes consisting of shale, for example, those on talus slopes above NN3 and NN4 (Fig. 5b). Many pebbly rock glaciers are situated below talus slopes abundant in fine materials derived from shale or limestone. Traces of debris flows remain on such finerich slopes (Fig. 5). In addition, a veneer of fine debris, originating from a small debris flow or snow avalanche, is often observed on late-lying snow banks covering the upper part of pebbly rock glaciers. In contrast, the latelying snow bank at the top of bouldery rock glaciers commonly shows isolated boulders, although it is occasionally covered by the deposit of a large rockfall (including a dirty snow avalanche).
287
represents a minimum of the original blocks, because the data include fragments shattered in situ after deposition. Thus, pebbly rock glaciers are distinguished from bouldery rock glaciers by the clast size of 15–20 cm in the mean b-axis diameter (Fig. 8a). The maximum b-axis diameter for each lithology is 350 cm (granite), 330 cm (massive limestone), 250 cm (gneiss and schist) and 89 cm (shale). The mean a and c-axes diameters of 25 clasts on pebbly rock glaciers are 8.9–19 cm and 2.5–5.3 cm, respectively, whereas those on bouldery rock glaciers are 34–155 cm and 13–68 cm, respectively. The mean a/c ratios of 25 clasts range from 2.7 to 11 (average 5.7) on pebbly rock glaciers and from 2.0 to 3.4 (average 2.8) on bouldery rock glaciers. The larger ratios for pebbly rock glaciers reflect more tabular clasts of shale. Clasts on the talus slopes above the rock glaciers are mainly cobbles and pebbles regardless of the type of rock glacier. The mean b-axis diameters of 25 clasts are 2.9–8.8 cm on pebbly rock glaciers, and those on bouldery rock glaciers are 6.5–20 cm (granite), 8.8– 18 cm (gneiss and schist) and 2.8–12 cm (massive limestone) (Fig. 8b). The average joint spacings of rockwalls are 9.2–11 cm (granite), 7.4–9.4 cm (gneiss and schist), 4.6–6.5 cm (massive limestone) and 2.0–
4.5. Surface composition Fig. 8 displays the downslope variation in clast size from the top of a talus slope to a rock glacier. The values at 0 m distance show the average joint spacing of the source rockwall. The mean b-axis (intermediateaxis) diameters of 25 clasts are 5.3–13 cm on pebbly rock glaciers mainly consisting of shale, whereas those on bouldery rock glaciers are 24–105 cm for granite, 21–58 cm for gneiss and schist and 24–55 cm for massive limestone. The value for massive limestone
Fig. 9. Stratigraphy of BNU rock glacier. Cobbles and pebbles filled with sandy–silty matrix were entirely frozen below 1 m deep on 3 August, 2000. The shaded area represents frozen debris. The pit did not reach the bottom of permafrost.
288
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
below 2 m depth. The gravimetric ice content of the borecores (c. 10 cm long) was 50% at 4 m depth and 28% at 5 m depth. Ignoring air voids, the volumetric ice content was estimated to be 60% at 4 m depth and 45% at 5 m depth. The borehole did not reach the bedrock at the bottom (5.4 m deep). In addition, five shallow excavations at BNU, BNL and BW revealed the thickness of the uppermost openwork layer varies from 3 cm (one or two fragments) to 50 cm.
3.2 cm (shale), which are comparable with the average clast sizes on the talus slopes (Fig. 8b). In general, the prevailing diameters of surface clasts range from 3 to 25 cm on pebbly rock glaciers (Table 1). Boulders are nearly absent on half of the pebbly rock glaciers. Six pebbly rock glaciers (A3, A4U, A4L, A5U, A5L, J1), derived from limestone, lack boulders and sandy–silty matrix near the surface. Boulders are, however, sparsely distributed on pebbly rock glaciers originating from rockwalls producing blocks and fine clasts (A6, A7, A8, BNL, BN3U, BN3L, BW1, NN3, NN10, NN11, NN12). In particular, BN2 and NN4 are covered with a thin openwork bouldery layer on a matrix-supported layer.
5.2. Seismic refraction soundings The shallow seismic refraction soundings indicated a two-layered structure in three bouldery and nine pebbly rock glaciers (Table 2, see also Fig. 5 for the locations). The first layer with a low P-wave velocity (330–650 m s 1) indicates the active layer and the second layer with a high P-wave velocity (2000–4400 m s 1) permafrost. The computed active layer is 2 to 3 m thick on many active and inactive rock glaciers, but thick active layers (N4 m) are indicated on NN8 and the lower part of BNL (BNL-b), BW2 and NN2. The latter implies degrading permafrost, although these rock glaciers lack a depressed transverse profile indicative of melting ice-supersaturated permafrost (cf. Ikeda and Matsuoka, 2002).
5. Internal structure 5.1. Direct observation The pit and borehole on BNU displayed that platy shale pebbles and cobbles filled with sandy–silty matrix are the major components, except for the uppermost 40 cm of an openwork layer (Fig. 9). The frost table lay at about 1 m depth on August 3, 2000. Below the frost table, the debris was entirely ice-cemented to the bottom of the borehole and lacking visible air voids. A number of small ice lenses (b2 cm thick) were observed
Table 2 Results of seismic refraction and vertical electrical resistivity soundings Site
P-wave stratigraphy First layer V(m s 1)
Second layer
AB
First layer
Second layer
Third layer
Fourth layer
AB/2
V(m s 1)
(m)
U(kgm)
D(m)
U(kgm)
U(kgm)
D(m)
U(kgm)
(m)
4100 3700 4400
55 60 60
19–21 6.8–56
2.4 1.3
260 5.6
4.4 4.2
520 210
17 17
10 1.3
100 100
2.2–2.6 2.1–2.3 4.4–5.8
2800 2900 3400
50 50 40
0.68 4.1–1.4
0.51 1.4
2.2 2.8
5.0 12
3.2 6.7
2.1–2.7 2.0–2.1 2.3–5.3 3.6–4.4 4.8–8.6
3000 2300 2000 2600 3200
42 40 55 45 60
2.2–2.4
2900
50
7.8–29 1.9 0.89 2.6 2.7 2.1–1.6 2.0–0.81 3.8–4.2 4.6–3.2
1.5 1.3 1.4 1.3 0.51 2.7 2.7 4.5 3.8
2.5 0.79 0.81 0.55 2.0 1.2 6.5 8.5 6.4
26 3.7 12 4.1 4.8 8.3 10 16 8.4
1.5 2.0 0.12 0.44 1.8 2.0 5.1 2.0 12
D(m)
Bouldery rock glaciers PN1 550 2.8–2.9 PN5 650 1.3–2.4 PS3* 330 2.6–3.1 Pebbly rock glaciers BNU 390 BNL-a 370 BNL-b 470 BN2** BN3U* 350 BW1 330 BW2 350 NN2 420 NN8* 450 NN10 NN11 NN12 340
DC resistivity stratigraphy
D(m)
20 31
23 22 15 14
23
0.66 0.32
1.1 21 1.8 2.1
0.022
50 100 50 50 80 50 50 50 50 50 100
*Partly vegetated (probably inactive). **A thin openwork bouldery layer on the surface. Symbols: V (P-wave velocity), D (depth of layer base), AB (length of sounding profile) and U (calculated resistivity).
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
5.3. Electrical DC resistivity soundings The vertical soundings showed that the DC resistivities in pebbly rock glaciers were usually one to two orders of magnitude lower than that of typical bouldery rock glaciers at a similar depth (Table 2, Fig. 10). Permafrost in PN5 and PS3 bouldery rock glaciers showed resistivity values of 210–520 kVm. In contrast, the resistivity values of the permafrost in 11 pebbly rock glaciers ranged from 0.44 to 8.5 kVm. In addition, the resistivities in typical bouldery rock glaciers significantly increase below the frost table estimated by the seismic soundings, whereas the values in pebbly rock glaciers only slightly change across the frost table. The sharp decrease in resistivity in BN2 reflects the different clast size between the uppermost bouldery layer and the underlying pebbly layer. The third layer of BNL
289
(6.7 kVm) and NN12 (12 kVm) may also show the resistivities of permafrost. The resistivity imaging displayed a spatial variation in resistivity within a rock glacier. In a longitudinal section, elliptic layers with relatively high resistivity are distinguished from the surroundings by a large resistivity gradient (Figs. 11 and 12). The observed resistivities show a similar range to those obtained from the vertical soundings, although the cores with relatively high resistivity in PN5 (N 1000 kVm) and BW1 (N2 kVm) were not detected by the vertical soundings (Figs. 11 and 12, see also Fig. 10). 5.4. Interpretation The large gradient in resistivity at shallow depth in bouldery rock glaciers certainly indicates the presence
Fig. 10. Apparent resistivity curves and calculated electrical DC resistivities. Parenthetic symbols: b = bouldery rock glacier; p = pebbly rock glacier; and pb = pebbly rock glacier covered with a thin openwork bouldery layer.
290
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Fig. 11. Electrical DC resistivity tomograms of bouldery rock glaciers.
of permafrost (e.g. PN5 in Fig. 10). This contrast is small or even absent in the pebbly rock glaciers, although the P-wave velocities indicate the presence of permafrost. The resistivity values reflect temperature and grain-size distribution (e.g. Hoekstra and McNeill, 1973). For example, the active layer of the bouldery rock glaciers (resistivity 10–50 kVm) has resistivity an order of magnitude higher than that of the pebbly rock glaciers (1–5 kVm), because the interstitial fine materials of the latter hold water that reduces the resistivity. Similarly, the low resistivity of permafrost in the pebbly rock glaciers reflects the higher content of fine debris compared with the bouldery rock glaciers. Permafrost temperature may also affect the resistivity. The observed rock glaciers, lying near the lower limit of mountain permafrost, probably have permafrost close to the melting point (0 to 28C: Arenson et al., 2002). In particular, the decreasing resistivities (BW1, BW2, NN2, NN8) may result from permafrost at the melting point, as suggested by the relatively warm location and/or thick active layer (Table 2, Fig. 10, see also Fig. 5 for the location). The two orders of magnitude difference in resistivity between bouldery and pebbly rock glaciers, however, is mainly attributed to the difference in ice content. The
previous measurements have reported a wide range of resistivity (5–2000 kVm), which probably reflects the variable ice contents from less than 60% in ice-cemented debris (e.g. Fisch et al., 1977; Barsch et al., 1979; Elconin and LaChapelle, 1997; Arenson et al., 2002) to nearly 100% in massive ice (e.g. Vonder Mu¨hll and Holub, 1992; Arenson et al., 2002). Thus, the high resistivity (N 100 kVm) in the bouldery rock glaciers probably results from a highly ice-supersaturated layer (Haeberli et al., 1998), whereas the lower range (b 20 kVm) in the pebbly rock glaciers mostly indicates the lack of massive ice. The two-dimensional images display that permafrost in bouldery rock glaciers have a large spatial variation in resistivity (100–4000 kVm in PN5, 10– 250 kVm in PS3), probably depending on permafrost temperature and ice content (Fig. 11). The high resistivities (N 100 kVm) prevailing under both PN5 and PS3 probably represent highly ice-supersaturated layers. In addition, the very high value (N 1000 kVm) in PN5 indicates even the presence of massive ice (N 10 m thick) at relatively low temperature, whereas the lower range (b30 kVm) at the frontal part of PS3 indicates relatively low ice content and/or high temperature. In comparison, pebbly rock glaciers show much smaller spatial variation in resistivity (Fig. 12). The elliptic layers with relatively high DC resis-
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
291
Fig. 12. Electrical DC resistivity tomograms of pebbly rock glaciers. The depth of the frost table was estimated by seismic soundings.
tivity (4–20 kVm) in pebbly rock glaciers may reflect slightly colder permafrost than the surroundings. The resistivity lower than 2 kVm in the upper part of BNU, the lower part of BNL and the both ends of BW1 probably indicates permafrost at the melting point or even an unfrozen layer. This interpretation is supported by the permafrost temperatures in the borehole on BNU which are nearly 0.08C throughout the year at 3, 4 and 5 m depths (Ikeda et al., 2003).
Also, the low P-wave velocity indicates that the lower part of BNL has a thick unfrozen layer. 6. Processes Fig. 13 compares the morphological characteristics and major geomorphological processes between pebbly and bouldery rock glaciers. The differences between the two depend primarily on lithology of the source rock-
292
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Fig. 13. A schematic diagram of rock control on rock glaciers.
walls. Such a rock control on rock glaciers is discussed in this section. 6.1. Debris input The primary difference between bouldery and pebbly rock glaciers lies in the material composition that inherits characteristics of the source rockwall. The average joint spacing of the rockwalls is much smaller than the dimensions of the clasts on the bouldery rock glaciers (Fig. 8). This indicates that the clasts mainly originate from large rockfalls reflecting the spacing of long and deep cracks. In contrast, the clast size of the pebbly rock glaciers approximates the average joint spacing of the rockwalls (Fig. 8). Thus, the pebbly rock glaciers are mainly fed by small-scale rockfalls. Such pebbly rock glaciers develop mainly below rockwalls closely spaced by joints (cf. Matsuoka and Ikeda, 2001). In particular, shale and phyllite rockwalls are most effective in producing fine materials that constitute the matrix of the active layer (cf. Wahrhaftig and Cox, 1959; Calkin et al., 1987). The variation in clast size from talus slopes to rock glaciers also implies the processes of debris transport. The clast size suddenly increases from the foot of the talus slopes to the head of the bouldery rock glaciers (Fig. 8). This probably results from fall sorting, thereby boulders preferentially accumulating on the foot of talus slopes (e.g. Haeberli et al., 1998). Sorting may be further promoted by sieving and downwashing of fine materials (e.g. Vick, 1981; Haeberli, 1985), eventually resulting in an openwork bouldery layer.
In contrast, the clast size of the pebbly rock glaciers is similar to that on the talus slopes, since cobbles or finer debris preferentially produced from the densely jointed rockwalls are unfavourable for gravitational sorting. Instead, the talus slopes abundant in fine materials are susceptible to debris flow and/or solifluction, which efficiently transport fine materials on the pebbly rock glaciers. Such a transport process is generally lacking on the talus slopes above the active bouldery rock glaciers. The length of rock glaciers is correlated with the size of the source rockwalls (Fig. 7). In general, high, large rockwalls can produce abundant debris feeding long bouldery rock glaciers. In contrast, low rockwalls constrain the downslope growth of pebbly rock glaciers, because they produce a much smaller amount of debris. Shale and platy limestone, however, constitute densely jointed rockwalls that experience more rapid weathering than sparsely jointed rockwalls (e.g. Matsuoka, 1990). Despite the high weathering rate, the small rockwall cannot produce a large amount of debris which supports a large rock glacier (Matsuoka and Ikeda, 2001). In addition, high rockwalls would have intensified debris supply onto bouldery rock glaciers during the past deglaciation (Ballantyne, 2002). 6.2. Permafrost formation A wide range of DC resistivity values, observed in permafrost below bouldery and pebbly rock glaciers, implies a variety of non-glacial ice formation in the rock glaciers. The low resistivity (1–10 kVm) in the pebbly rock glaciers indicates ice-cemented to slightly
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
ice-supersaturated structure (Table 1, Figs. 10 and 12). The direct observation revealed a low resistivity (2 kVm) layer representing frozen debris cemented with ice (Figs. 9 and 12). Such a structure is basically formed by groundwater freezing accompanied by frost heave. Freezing of unconfined groundwater is unlikely to form a highly ice-supersaturated layer or even massive ice (resistivityN 100 kVm) in bouldery rock glaciers (see Fig. 11), which is indicated by the relatively icepoor structure of pebbly rock glaciers. One of the origins for the intensive ice-supersaturation in talusderived rock glaciers is burying snow banks. Longlasting or perennial snow banks, often covered with rockfall deposits, occur at the top of talus-derived rock glaciers (Dyke, 1990; Matsuoka, 1998). Such a debris accumulation can incorporate the snow bank into permafrost, producing an ice-supersaturated layer (e.g. Outcalt and Benedict, 1965; Haeberli et al., 1998; Berthling et al., 2000). The difference in ice content between bouldery and pebbly rock glaciers appears to reflect the different debris accumulation. The large rockwalls above bouldery rock glaciers have potential to produce a thick debris deposit on the snow bank through a large rockfall. The low rockwalls above pebbly rock glaciers are much less favourable for such a large rockfall. Pebbly rock glaciers below a high rockwall may have an ice-supersaturated layer, as indicated by a high resistivity layer (N 10 kVm) in the upper part of NN12 (see Fig. 12). Another origin of the ice-supersaturation is ice injection, which is the major mechanism of pingos and frost blisters (e.g. Mackey, 1973; van Everdingen, 1978; French, 1996). The presence of intrapermafrost water in rock glaciers (e.g. Vonder Mu¨hll, 1992; Vonder Mu¨hll et al., 2003) cannot rule out this possibility, but this process fails to explain the lack of massive ice in pebbly rock glaciers. 6.3. Permafrost deformation Most of the pebbly rock glaciers have a convex-up or flat longitudinal profile, thereby the surface gradient increases or maintains towards the front (Fig. 6). The increasing surface gradient results in the predominance of extending flow, which is responsible for the lack of transverse ridge–furrow topography produced by compressive flow (Haeberli et al., 1979). In contrast, many bouldery rock glaciers decline towards the front, where the ridge–furrow topography is common. The difference in the surface topography between bouldery and pebbly rock glaciers depends mainly on the length and location. Bouldery rock glaciers often develop into the
293
valley bottom where the surface gradient declines, whereas pebbly rock glaciers usually terminate within (or at the foot of) a valley-side slope. The exceptionally long pebbly rock glaciers developing along the valley bottom (A6, A7, A8) have the ridge-furrow topography (Fig. 2d). The longitudinal compression causes thickening of rock glaciers. The thickened ice-supersaturated layer results in increasing volumetric ice content. Thus, the ice-supersaturation is probably promoted by long-lasting compression, which is common in long bouldery rock glaciers. Such thickening explains also why many bouldery rock glaciers have higher frontal slopes (N 20 m) than pebbly rock glaciers. 6.4. Permafrost preservation Near the lower limit of mountain permafrost, the active layer composition also affects the preservation of permafrost, which controls the altitudes of the terminus of rock glaciers. The openwork structure of the active layer of bouldery rock glaciers permits intensive cooling of the underlying permafrost and the storage of cold air in winter (e.g. Keller and Gubler, 1993; Hoelzle et al., 1999; Matsuoka and Ikeda, 2003). Below a thick snow cover the surface of active pebbly rock glaciers maintain higher temperatures than that of active bouldery rock glaciers (Ikeda et al., 2003). Thus, bouldery rock glaciers have a potential to extend to lower altitudes. The frontal altitude of two active bouldery rock glaciers (PN1, PN2), 2540 m ASL, is lower than that of a nearby relict pebbly rock glacier (BN3L), 2610 m ASL (Fig. 5a). 6.5. Active layer modification The mobility of the active layer of rock glaciers depends on the presence of fine materials. Matrix-free blocks on bouldery rock glaciers move passively by creep of the underlying permafrost. The sharp frontal edge is maintained on active bouldery rock glaciers, because individual boulders in the active layer only fall or slide on the frontal slope. In contrast, rounded frontal edge of pebbly rock glaciers indicates solifluction or other soil transport in the matrix-supported layer. In addition, solifluction lobes develop from the side of a pebbly rock glacier in the Furggenta¨lti, Swiss Alps (Krummenacher and Budmigner, 1992), and patterned ground occurs on some pebbly rock glaciers in the Brooks Range, Alaska (Calkin et al., 1987). Such a soil movement may gradually decrease the thickness of pebbly rock glaciers.
294
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
7. Conclusions
Acknowledgements
The difference in the source rockwall geology, which controls the clast size, determines the two types of rock glaciers. Bouldery rock glaciers have the active layer composed of matrix-free boulders originating from resistant rocks (e.g. crystalline rocks and massive limestone), whereas pebbly rock glaciers consist mainly of pebbles and cobbles filled with sandy–silty matrix from less resistant rocks (e.g. shale and platy limestone). Pebbly rock glaciers are distinguished from bouldery rock glaciers by the clast size of 15–20 cm in the mean b-axis diameter. The clast size controls transport processes from the source to the front of a rock glacier, resulting in different shape, size and internal structure between pebbly and bouldery rock glaciers. The matrix-free boulders (20–100 cm in the mean baxis diameter) imply that large rockfalls accompanied by fall sorting of debris are the major process of the debris input on bouldery rock glaciers. In contrast, the pebbles and cobbles (5–10 cm in the mean b-axis diameter) filled with sandy–silty matrix imply that small-scale rockfalls, aided by debris flow and/or solifluction, feed the pebbly rock glaciers. The fine materials keep the active layer of pebbly rock glaciers susceptible to solifluction, which results in the rounded frontal edge. Seventy-five percent of the pebbly rock glaciers are shorter than 200 m, whereas bouldery rock glaciers are typically longer than 200 m. The source rockwalls of such short pebbly rock glaciers are lower than 50 m, whereas those of bouldery rock glaciers are generally higher than 100 m. The small exposure of the less resistant source rockwall strongly constrains debris supply. As a result, pebbly rock glaciers usually terminate within a valley-side slope, whereas many bouldery rock glaciers extend into the valley bottom. The location of pebbly rock glaciers results in the lack of compressive flow, which is responsible for the low frontal slope (V 20 m) and the lack of transverse ridge–furrow topography. The pit–borehole stratigraphy and low DC resistivities (b10 kVm) indicate ice-cemented or slightly supersaturated permafrost in the pebbly rock glaciers. Such a structure may largely originate from groundwater freezing. In contrast, highly ice-supersaturated structure (or massive ice) indicated by high DC resistivities (N100 kVm) in bouldery rock glaciers appears to originate from buried snow banks, which have been buried with large rockfall deposits from the large source rockwall.
This study was financially supported by the Grantin-Aid (No. 15500676) for Scientific Research and Grant-in-Aid for JSPS Fellows from the Ministry of Education, Science and Culture, Japan and by the Sasakawa Scientific Research Grant from the Japan Scientific Society. We acknowledge F. Keller for logistical help, K. Hirakawa, T. Watanabe, M. Aoyama, K. Fukui, Y. Sawada and S. Iwasaki for their cooperation in the field, and Mi. Abe, Ma. Abe, T. Date and Y. Eda for field assistance. References Arenson, L., Hoelzle, M., Springman, S., 2002. Borehole deformation measurements and internal structures of some rock glaciers in Switzerland. Permafr. Periglac. Process. 13, 117 – 135. Ballantyne, C.K., 2002. Paraglacial geomorphology. Quat. Sci. Rev. 21, 1935 – 2017. Barsch, D., 1977. Nature and importance of mass-wasting by rock glaciers in alpine permafrost environments. Earth Surf. Processes 2, 231 – 245. Barsch, D., Fierz, H., Haeberli, W., 1979. Shallow core drilling and bore-hole measurements in the permafrost of an active rock glacier near the Grubengletscher Wallis, Swiss Alps. Arct. Alp. Res. 11, 215 – 228. Berthling, I., Etzelmu¨ller, B., Isaksen, K., Sollid, J.L., 2000. Rock glaciers on Prins Karls Forland: II. GPR soundings and the development of internal structures. Permafr. Periglac. Process. 11, 357 – 369. Burger, K.C., Degenhardt Jr., J.J., Giardino, J.R., 1999. Engineering geomorphology of rock glaciers. Geomorphology 31, 93 – 132. Calkin, P.E., Haworth, L.A., Ellis, J.M., 1987. Rock Glaciers of Central Brooks Range, Alaska, U.S.A.. In: Giardino, J.R., Shroder Jr., J.F., Vitek, J.D. (Eds.), Rock Glaciers. Allen and Unwin, London, pp. 65 – 82. Chueca, J., 1992. A statistical analysis of the spatial distribution of rock glaciers, Spanish central pyrenees. Permafr. Periglac. Process. 3, 261 – 265. Dyke, A.S., 1990. A lichenometric study of Holocene rock glaciers and neoglacial moraines, Frances Lake Map Area, southeastern Yukon Territory and Northwest Territories. Pap. - Geol. Surv. Can. 394, 1 – 33. Elconin, R.F., LaChapelle, E.R., 1997. Flow and internal structure of a rock glacier. J. Glaciol. 43, 238 – 244. Evin, M., 1987. Lithology and fracturing control of rock glaciers in southwestern Alps of France and Italy. In: Giardino, J.R., Shroder Jr., J.F., Vitek, J.D. (Eds.), Rock Glaciers. Allen and Unwin, London, pp. 83 – 106. Fisch Sr., W., Fisch Jr., W., Haeberli, W., 1977. Electrical D.C. resistivity soundings with long profiles on rock glaciers and moraines in the Alps of Switzerland. Z. Gletsch.kd. Glazialgeol. 13, 239 – 260. Frauenfelder, R., Ka¨a¨b, A., 2000. Towards a paleoclimatic model of rock glacier formation in the Swiss Alps. Ann. Glaciol. 31, 281 – 286. Frauenfelder, R., Haeberli, W., Hoelzle, M., Maisch, M., 2001. Using relict rockglaciers in GIS-based modelling to reconstruct Younger
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296 Dryas permafrost distribution patterns in the Err–Julier area Swiss Alps. Nor. Geogr. Tidsskr. 55, 195 – 202. Frauenfelder, R., Haeberli, W., Hoelzle, M., 2003. Rockglacier occurrence and related terrain parameters in a study area of the Eastern Swiss Alps. Proc. 8th Int. Conf. Permafrost. Balkema, Lisse, Netherlands, pp. 253 – 258. French, H.M., 1996. The Periglacial Environment, 2nd ed. Longman, London. Frich, P., Brandt, E., 1985. Holocene talus accumulation rates, and their influence on rock glacier growth. A case study from Igpik, Disko, West Greenland. Geogr. Tidsskr. 85, 32 – 44. Giradino, J.R., Vitek, J.D., 1988. The significance of rock glaciers in the glacial–periglacial landscape continuum. J. Quat. Sci. 3, 97 – 103. Haeberli, W., 1985. Creep of mountain permafrost: internal structure and flow of alpine rock glaciers. Mitt. Versuchsanst. Wasserb., Hydrol. Glaziol. ETH Zu¨r. 77 (142 pp.). Haeberli, W., Patzelt, G., 1982. Permafrostkartierung im Gebiet der ¨ tztaler Alpen. Z. Hochebenkar–Blockgletscher, Obergurgl, O Gletsch.kd. Glazialgeol. 18, 127 – 150. Haeberli, W., Vonder Mu¨hll, D., 1996. On the characteristics and possible origins of ice in rock glacier permafrost. Z. Geomorphol. N. F. Suppl.Bd. 104, 43 – 57. Haeberli, W., King, L., Flotron, A., 1979. Surface movement and lichen-cover studies at the active rock glacier near the Grubengletscher, Wallis, Swiss Alps. Arct. Alp. Res. 11, 421 – 441. Haeberli, W., Evin, M., Tenthorey, G., Kensen, H.R., Hoelzle, M., Keller, F., Vonder Mu¨hll, D., Wagner, S., Pelfini, M., Smiraglia, C., 1992. Permafrost research sites in the Alps: excursions of the international workshop on permafrost and periglacial environments in mountain areas. Permafr. Periglac. Process. 3, 189 – 202. Haeberli, W., Hoelzle, M., Keller, F., Vonder Mu¨hll, D., Wagner, S., 1998. Ten years after the drilling through the permafrost of the active rock glacier Murte`l, eastern Swiss Alps: answered questions and new perspectives. Proc. 7th Int. Conf. Permafrost. Centre d’e´tudes Nordiques, Sainte-Foy, Canada, pp. 403 – 410. Haeberli, W., Ka¨a¨b, A., Wagner, S., Vonder Mu¨hll, D., Geissler, P., Haas, J.N., Glatzel-Mattheier, H., Wagenbach, D., 1999. Pollen analysis and 14C age of moss remains in a permafrost core recovered from the active rock glacier Murte`l–Corvatsch, Swiss Alps: geomorphological and glaciological implications. J. Glaciol. 45, 1 – 8. Hoekstra, P., McNeill, D., 1973. Electromagnetic probing of permafrost. Proc. 2nd Int. Conf. Permafrost. National Academy of Sciences, Washington, pp. 517 – 526. Hoelzle, M., 1998. Rock glaciers, Upper Engadin, Switzerland. IPA Data and Information Working Group, Comp. Circumpolar Active-layer Permafrost System (CAPS), ver. 1.0. National Snow and Ice Data Center, Boulder, CD-ROM, ROM. Hoelzle, M., Wegmann, M., Krummenacher, B., 1999. Miniature temperature dataloggers for mapping and monitoring of permafrost in high mountain areas: first experience from the Swiss Alps. Permafr. Periglac. Process. 10, 113 – 124. Humlum, O., 2000. The geomorphic significance of rock glaciers: estimates of rock glacier debris volumes and headwall recession rates in West Greenland. Geomorphology 35, 41 – 67. Hunter, J.A.M., 1973. The application of shallow seismic methods to mapping of frozen surficial materials. Proc. 2nd Int. Conf. Permafrost. National Academy of Sciences, Washington, pp. 527 – 535.
295
Ikeda, A., Matsuoka, N., 2002. Degradation of talus-derived rock glaciers in the Upper Engadin, Swiss Alps. Permafr. Periglac. Process. 13, 145 – 161. Ikeda, A., Matsuoka, N., Ka¨a¨b, A., 2003. A rapidly moving small rock glacier at the lower limit of the mountain permafrost belt in the Swiss Alps. Proc. 8th Int. Conf. Permafrost. Balkema, Lisse, Netherlands, pp. 455 – 460. Ishikawa, M., Watanabe, T., Nakamura, N., 2001. Genetic differences of rock glaciers and the discontinuous mountain permafrost zone in Kanchanjunga Himal, eastern Nepal. Permafr. Periglac. Process. 12, 243 – 253. Ka¨a¨b, A., Gudmundsson, G.H., Hoelzle, M., 1998. Surface deformation of creeping mountain permafrost. Photogrammetric investigations on Murte`l rock glacier, Swiss Alps. Proc. 7th Int. Conf. Permafrost. Centre dTe´tudes Nordiques, Sainte-Foy, Canada, pp. 531 – 537. Keller, F., Gubler, H., 1993. Interaction between snow cover and high mountain permafrost Murte`l/Corvatsch, Swiss Alps. Proc. 6th Int. Conf. Permafrost, vol. 1. South China University Technology Press, Wushan, China, pp. 332 – 337. Krummenacher, B., Budmigner, K., 1992. Monitoring of periglacial phenomena in the Furggenta¨lti Swiss Alps. Permafr. Periglac. Process. 3, 149 – 155. Mackey, J.R., 1973. The growth of pingos, western Arctic coast, Canada. Can. J. Earth Sci. 10, 979 – 1004. Matsuoka, N., 1990. The rate of bedrock weathering by frost action: field measurements and a predictive model. Earth Surf. Process. Landf. 15, 73 – 90. Matsuoka, N., 1998. Monitoring rockwall instability in the Murtel–Corvatsch region, Upper Engadin. Beitra¨ge aus der Gebirgs-Geomorphologie, Mitteilungen der Versuchsanstalt fu¨r Wasserbau, Hydrologie und Glaziologie ETH Zu¨rich, vol. 158, pp. 7 – 14. Matsuoka, N., Sakai, H., 1999. Rockfall activity from an alpine cliff during thawing periods. Geomorphology 28, 309 – 328. Matsuoka, N., Ikeda, A., 2001. Geological control on the distribution and characteristics of talus-derived rock glaciers. Annu. Rep. Inst. Geosci., Univ. Tsukuba 27, 11 – 16. Matsuoka, N., Ikeda, A., 2003. Intensive cooling through an open-work blocky layer: ground temperature monitoring in the Swiss Alps. Annu. Rep. Inst. Geosci., Univ. Tsukuba 29, 21 – 25. Matsuoka, N., Hirakawa, K., Watanabe, T., Moriwaki, K., 1997. Monitoring of periglacial slope processes in the Swiss Alps: the first two years of frost shattering, heave and creep. Permafr. Periglac. Process. 8, 155 – 177. Matsuoka, N., Ikeda, A., Date, T., 2005. Morphometric analysis of solifluction lobes and rock glaciers in the Swiss Alps. Permafr. Periglac. Process. 16, 99 – 113. Outcalt, S.I., Benedict, J.B., 1965. Photo-interpretation of two types of rock glaciers in the Colorado Front Range, U.S.A. J. Glaciol. 5, 849 – 856. Palmer, D., 1986. Refraction Seismics. Geophysical Press, London. van Everdingen, R.O., 1978. Frost mounds on Bear Rock, near Fort Norman, N.W.T. 1975-1976. Can. J. Earth Sci. 15, 263 – 276. Vick, S.G., 1981. Morphology and the role of landsliding in formation of some rock glaciers in the Mosquito Range, Colorado. Geol. Soc. Amer. Bull. 92, 75 – 84. Vonder Mu¨hll, D.S., 1992. Evidence of intrapermafrost groundwater flow beneath an active rock glacier in the Swiss Alps. Permafr. Periglac. Process. 3, 169 – 173.
296
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Vonder Mu¨hll, D.S., Holub, P., 1992. Borehole logging in alpine permafrost, Upper Engadin, Swiss Alps. Permafr. Periglac. Process. 3, 125 – 132. Vonder Mu¨hll, D.S., Hauck, C., Lehmann, F., 2000. Verification of geophysical models in Alpine permafrost by borehole information. Ann. Glaciol. 31, 300 – 306.
Vonder Mu¨hll, D.S., Arenson, L.U., Springman, S.M., 2003. Temperature conditions in two alpine rock glaciers. Proc. 8th Int. Conf. Permafrost. Balkema, Lisse, Netherlands, pp. 1195 – 2000. Wahrhaftig, C., Cox, A., 1959. Rock glaciers in the Alaska Range. Bull. Geol. Soc. Am. 70, 383 – 436.
Pebbly versus bouldery rock glaciers: Morphology, structure and processes Atsushi Ikeda a,b,*, Norikazu Matsuoka c a
b
Division of Earth Sciences, National Institute of Polar Research, Itabashi, Tokyo 173-8515, Japan Water and Environmental Research Center, Institute of Northern Engineering, University of Alaska-Fairbanks, Fairbanks, Alaska 99775-5860, USA c Graduate School of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8572, Japan Received 20 January 2005; received in revised form 21 July 2005; accepted 21 July 2005 Available online 26 September 2005
Abstract Differences in rockwall geology cause two types of rock glaciers: a bouldery rock glacier, having the active layer composed of matrix-free boulders derived from crystalline rocks and massive limestone; and a pebbly rock glacier, consisting of matrix-supported debris derived from less resistant shale and platy limestone. Such material composition controls transport processes responsible for the shape, size and internal structure of the two types of rock glaciers. This paper compares the major processes controlling the morphology and structure of the two types in the Swiss Alps, based on mapping, description of geological and morphological characteristics, direct observation of stratigraphy and geophysical soundings of internal structure. In the Swiss Alps, pebbly rock glaciers are distinguished from bouldery rock glaciers by the clast size of 15–20 cm in the mean b-axis diameter. The former are fed by small-scale rockfalls, debris flows and solifluction, whereas the latter mainly originate from large rockfalls. Pebbly rock glaciers are generally smaller (b 200 m in length) than bouldery rock glaciers, because the small exposure of the less resistant source rockwall (b50 m in height) strongly constrains debris supply. As a result, pebbly rock glaciers usually terminate within a valley-side slope, whereas many bouldery rock glaciers extend into the valley bottom. The location of pebbly rock glaciers results in the low frontal slope (V 20 m) and the lack of transverse ridge-furrow topography, because of the lack of compressive flow. The pit-borehole stratigraphy and low DC resistivities (b 10 kVm) indicate ice-cemented or slightly supersaturated permafrost in the pebbly rock glaciers, which presumably originates from groundwater freezing. In contrast, highly ice-supersaturated structure indicated by high DC resistivities (N 100 kVm) in bouldery rock glaciers appears to originate from snow banks buried with deposits of large rockfalls from the large source rockwall. D 2005 Elsevier B.V. All rights reserved. Keywords: Rock glacier; Mountain permafrost; Debris supply; Geophysical soundings; Rock control; Swiss Alps
1. Introduction
* Corresponding author. Division of Earth Sciences, National Institute of Polar Research, Itabashi, Tokyo 173-8515, Japan. Tel.: +81 3 3962 4695; fax: +81 3 3962 5741. E-mail address: [email protected] (A. Ikeda). 0169-555X/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.geomorph.2005.07.015
The dynamics of rock glaciers reflects geology of the source rockwall, which controls the manner of debris supply. A few studies have highlighted the geological constraint on the development of rock glaciers (Wahrhaftig and Cox, 1959; Evin, 1987; Chueca,
280
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Fig. 1. Typical bouldery rock glaciers in the study area. (a) Murte`l I (MT) rock glacier having well-developed ridges and furrows and rimmed with a steep frontal slope 20 m in height. The surface boulders consist of gneiss and schist. (b) PN4 and PN5 rock glaciers originating from a massive limestone rockwall. The total height of the front reaches about 50 m.
1992). These studies state that rock glaciers mainly develop in lithology producing blocky debris, whereas they are infrequent on slopes composed of platy debris with finer materials. The size of clasts composing rock glaciers varies significantly with lithology. Evin (1987) has found that most rock glaciers in the southwestern Alps consist of coarse blocky debris derived from resistant rockwalls (granite, gneiss, sandstone and massive limestone), whereas rock glaciers derived from less resistant schistose rockwalls contain a large
amount of fine debris and small pore volume. Other studies have examined the volumetric relationships between rockwalls and rock glaciers (e.g. Barsch, 1977; Frich and Brandt, 1985; Humlum, 2000). Details are rarely understood, however, on the processes that the source rockwall contributes to the dynamics of rock glaciers. Here, we focus on the material composition of rock glaciers, which reflects the lithology of the source rockwall and controls slope processes. In this respect, we
Fig. 2. Pebbly rock glaciers in the study area. (a) BN rock glacier consists of the upper lobe (BNU) and lower lobe (BNL). The surface clasts are mainly shale pebbles and cobbles. BNL lacks a steep frontal slope. (b) NN10-12 rock glaciers, consisting mainly of shale pebbles and cobbles. (c) NN2 rock glaciers lying on the middle of a talus slope. (d) Exceptionally long (740 m) pebbly rock glacier A7, flowing along a valley bottom.
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Fig. 3. The location of the study area.
classify rock glaciers in terms of composition of surface materials into two types (Matsuoka et al., 2005). A bouldery rock glacier has an active layer composed of matrix-free boulders (Fig. 1). Most of the previously reported rock glaciers belong to this type. In contrast, a pebbly rock glacier is covered with pebbles and cobbles, mostly supported by sandy and silty matrix between coarse materials (Fig. 2). The latter type is probably comparable to the earthy rock glacier defined by Evin (1987). This paper compares the two types of rock glaciers in the Swiss Alps in terms of morphology and structure.
281
Field investigations involved mapping, description of geological and morphological characteristics, excavation with portable drills, seismic refraction soundings and one- and two-dimensional electrical resistivity soundings. These investigations mainly featured pebbly rock glaciers that have been poorly described, whereas morphology and structure of bouldery rock glaciers in the study area have already been well documented (e.g. Ka¨a¨b et al., 1998; Frauenfelder and Ka¨a¨b, 2000; Ikeda and Matsuoka, 2002). Ice is another vital component of rock glaciers, although its origin is problematic. In the case where a large rock glacier develops from a cirque floor or the head of a U-shaped valley, glacial, periglacial and their continuum origins have been proposed (e.g. Outcalt and Benedict, 1965; Giradino and Vitek, 1988; Humlum, 2000). In this paper, subsurface perennial ice in rock glaciers is called permafrost regardless of the origin because we focus on talus-derived rock glaciers that have had no contact with glaciers during the Holocene (cf. Frauenfelder et al., 2001). The development of nonglacial ice in rock glaciers is also discussed by comparing the structure and processes of pebbly and bouldery rock glaciers.
Fig. 4. Contour map (a) and geological map (b) of the study area, shown with the distribution of rock glaciers.
282
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
2. Study area Field investigations were undertaken in the Upper Engadin in the Swiss Alps, where a large extent of high mountain slopes permits the widespread occurrence of permafrost (Figs. 3 and 4a). The timberline lies at 2000–2200 m ASL, the mean annual air temperature 08C isotherm at 2200–2300 m ASL and the glacial equilibrium line at about 3000 m ASL (Haeberli et al., 1992). In contrast with the limited extent of glaciers, rock glaciers dominate mountain areas with peaks around 3000 m ASL (e.g. Frauenfelder et al., 2001; Ikeda and Matsuoka, 2002). About 80% of the active rock glaciers are considered to be talus-derived (Frauenfelder et al., 2003). Glacial erosion during the Pleistocene basically determined the morphology of mountain slopes in the study area. Bedrock lithology appears to primarily con-
trol the resistance of the slopes to glacial and paraglacial erosion. The bedrock in this region is broadly divided into the crystalline rock group (granite, granodiorite, diorite, gneiss and schist) and the sedimentary rock group (limestone, dolomite, conglomerate and shale) (Fig. 4b). The crystalline rocks generally sustain higher rockwalls than the sedimentary rocks (Matsuoka and Ikeda, 2001). The bedrock lithology also affects the size of debris: massive limestone, conglomerate and most of the crystalline rocks produce coarse debris, whereas platy/porous limestone and shale favour production of fine debris. The difference in size of the debris affects periglacial landforms. In general, coarse debris constructs talus slopes and rock glaciers below high rockwalls, whereas fine debris promotes development of solifluction lobes and pattern ground on gentle slopes (Matsuoka et al., 1997) and, where debris sedimentation is sufficient, can produce rock glaciers.
Fig. 5. Distribution of rockwalls and rock glaciers in the Corviglia region. (a) Northern area. (b) Southern area. Small-scale topography and the sites of geophysical soundings are also indicated.
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
The major study area is Corviglia, a large ski region north of St. Moritz (Fig. 5). In the northern Corviglia area, granite and limestone are the major lithologies composing rockwalls and shale is partly exposed on the flank of Piz Padella and Las Trais Fluors (Fig. 5a). In the southern Corviglia area, conglomerate composes the southern face of Piz Nair, whereas shale partly embedding chert composes the northern face (Fig. 5b). The high shale rockwall of Piz Nair is supported by the overlying conglomerate caprock. The morphology of pebbly rock glaciers were also studied on the northern face of Piz Blaisun, lying north of Albulapass, and the northern area of Julierpass (Fig. 4). The former area is mainly composed of platy limestone, and the latter massive limestone. Both areas also include shale outcrops. 3. Methodology 3.1. Distribution and morphology Bouldery and pebbly rock glaciers were mapped on the 1/25,000 topographic maps of Switzerland on the basis of field and air photograph surveys. Previous information on the distribution of rock glaciers in the Upper Engadin by Hoelzle (1998) and Frauenfelder et al. (2001) were also incorporated in the mapping. The lithologies of these rock glaciers were determined in the field with the help of the 1/200,000 geological map of Engadin. Small-scale topography, such as ridges and furrows, lobate bulges and debris flow levees on and around rock glaciers, was also mapped on the 1/15,000 farmland-name map of Celerina, which covers the Corviglia area. Morphological parameters of 28 pebbly rock glaciers were measured on the maps and in the field. The parameters include the aspect, frontal altitude, length, width, frontal height, surface gradient and maximum height of headwall. The frontal height was defined as the height of the steep frontal slope distinguished from the upper surface by a nick-point. In the field, longitudinal profiles of two bouldery and 18 pebbly rock glaciers were determined with a clinometer and an ultrasonic distance meter. 3.2. Surface composition The prevailing clast size was evaluated on 25 pebbly rock glaciers. In addition, the size of the surface clasts was more accurately determined on ten rock glaciers and talus slopes along the longitudinal profile. The measurement sites included four pebbly rock glaciers
283
consisting of shale (BNU, BNL, BW2, NN8) and four bouldery rock glaciers consisting of granite (O1, O2) or massive limestone (PN5, PS3) in the Corviglia area (see Figs. 4a and 5 for the locations). The measurement was also carried out on two rock glaciers consisting of gneiss and schist near the Corviglia area: Murte`l (MT) rock glacier lying 8 km south of St. Moritz and Muragl (MR) rock glacier lying 6 km east of St. Moritz, where intensive observations have been undertaken (e.g. Haeberli et al., 1999; Arenson et al., 2002). On each profile, four to eight horizontal survey lines were chosen. These lines were located longitudinally at 18– 130 m intervals. For each line, 25 clasts were sampled at intervals of 0.5, 1 or 2 m and the three-axis diameters were measured. The interval depended on the predominant size of the clasts. The size of rockfall debris is controlled by the joint spacing of the source rockwall (Matsuoka and Sakai, 1999). Average joint spacing was also measured at the foot of rockwalls above the talus slope. The number of joints, crossing a ring wire 2 m in circumference, was counted on the rock face and divided by 2 m to compute the average joint spacing. The next measurement was made by shifting the wire by 0.6–0.7 m, and in total 10 measurements were performed at one site. Three sites were chosen on a rockwall. As a result, the average joint spacing on the rockwall represented the mean value for 30 points. 3.3. Internal structure 3.3.1. Direct observation The internal structure of the BNU pebbly rock glacier was directly observed by excavation and drilling in early August 2000. A 3-m deep pit was excavated with a shovel and rock-cutting machine and then an additional 2.4 m deep borehole (6 cm in diameter) was drilled with an engine auger down from the bottom of the pit. The stratigraphy was observed on the pit wall and with the borecores. The gravimetric ice content was determined for the borecores at 4 and 5 m depths. 3.3.2. Seismic refraction soundings The presence of permafrost was investigated in eight pebbly and three bouldery rock glaciers by seismic refraction soundings. A high velocity layer (z 2000 m s 1) is regarded to be frozen where it underlies a low velocity layer (b 1500 m s 1) much thinner than the apparent thickness of the rock glacier (e.g. Hunter, 1973; Haeberli and Patzelt, 1982). A 3-channel seismograph, McSEIS-3 (manufactured by Oyo, Japan) was used for the survey. A 4-kg sledge-
284
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
permafrost and/or between layers having different grain size distribution, DC resistivities reinforce the interpretation of the internal structure of rock glaciers suggested by borecores or other geophysical properties (e.g. Fisch et al., 1977; Haeberli and Vonder Mu¨hll, 1996; Vonder Mu¨hll et al., 2000). In particular, the resistivity value is an indicator of unfrozen water content in permafrost, which mainly reflects ground temperature and grain-size distribution. One-dimensional (vertical) electrical DC resistivity soundings were performed on two bouldery and eleven pebbly rock glaciers from late July to early August in 1999 and 2002 with the resistivity meter SYSCAL R1 PLUS (manufactured by Iris Instruments, France). The setting of the electrodes followed the Schlumberger array. The resistivity values of the subsurface layers and the boundaries between these layers were calculated with the software ElecAuto (Oyo, Japan).
hammer produced a seismic pulse. The shot points were placed at 2 or 2.5 m intervals near the receivers and otherwise at 5 or 6 m intervals. A survey line was 40– 60 m long. The P-wave velocities of two or three layers and the depth of the layer boundaries were determined after a correction with regard to the dip of the boundary (Palmer, 1986). The soundings were carried out from middle July to middle August in 2000–2003. 3.3.3. DC resistivity soundings The structure in rock glaciers was investigated by direct current (DC) resistivity soundings. Resistivity in ground increases when ground temperature falls below 0 8C because of a decrease in liquid water content (e.g. Hoekstra and McNeill, 1973). Thus, if water content is homogeneous in the ground, permafrost is detected by DC resistivity soundings. Although the water content varies between the active layer and Table 1 Morphology and composition of pebbly rock glaciers Site
Vegetation1
A1U A1L A3 A4U A4L A5U A5L A6 A7 A8 BNU BNL BN2 BN3U BN3L BW1 BW2 NN1 NN2 NN3 NN4 NN8 NN9 NN10 NN11 NN12 J1 J2
NV NV NV NV NV NV NV NV NV NV NV NV NV PV DV NV NV DV NV PV PV PV DV NV NV NV NV NV
Slope aspect
Frontal altitude
Length
Width
Frontal slope ht.
Surface gradient
(8)
(m)
(m)
(m)
(m)
(8)
65 65 30 355 355 15 15 290 275 336 35 35 40 340 350 270 270 355 5 5 10 35 70 35 20 30 90 275
2570 2540 2540 2645 2615 2650 2605 2670 2550 2385 2800 2765 2730 2640 2610 2790 2810 2570 2605 2620 2640 2705 2710 2735 2755 2770 2835 2735
45 75 55 35 55 55 85 505 740 500 50 100 45 90 50 110 90 35 90 55 140 240 80 245 290 225 40 90
110 70 100 85 70 70 110 160 140 180 130 105 120 230 165 100 90 100 65 110 130 100 120 100 90 200 180 40
16 4 9 34 9 11 14 20 25 20 10 0 20 4 3 5 13 9 14 8 20 10 5 7 na* na* 10 3
23 21 18 24 25 23 21 15 17 21 25 20 18 28 18 23 28 13 30 21 17 17 19 20 15 18 15 29
Lithology2
Dominant clast size
Location3
(cm) LSp, SH LSp, SH LSp LSp LSp LSp LSp LSp, LSm LSp, LSm LSp, LSm SH, LSm SH, LSm LS m, SH SH, LSm SH, LSm SH, LSm SH SH SH, CG SH, CG CG, SH SH SH SH, CG SH, CG SH, CG LSp SH
5–10 5–10 10–20 10–20 10–20 10–20 10–20 15–25 15–25 5–10 5–15 N20 5–10 5–10 3–8 3–8 3–8 5–10 N20 5–10 5–15 5–15 5–15 10–20 5–10
Headwall max. height
Other features4
(m) TM TF TF TM TM TM TF VF VF VF TM TF TF TM TF TF TF TF TM TF TF TF VF TF TF TF TF TM
b5 – – b5 – 40 – 200 340 320 30y – 40 10y – 30y b5 50 100 50 20 150 – 160 190 160 10 b5
OW OW OW OW OW RF RF RF
OWB RF
RF
OWB RF
OW
1 NV = not or sparsely vegetated (probably active); PV = partly vegetated (probably inactive); DV = densely vegetated (probably relict). 2 LSp = platy limestone/dolomite; LSm = massive limestone/dolomite; SH = shale; CG = conglomerate. 3 TF = foot of a talus slope; TM = middle of a talus slope; VF = valley floor. 4 OW = neither boulder nor sandy–silty matrix near the surface; OWB = a thin openwork bouldery layer on the surface; RF = transverse ridges and furrows on the surface. *Artificially modified front. yThe height of a caprock consisting of massive limestone: the underlying shale rockwall is lower than 5 m.
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
The two-dimensional internal structure of two bouldery and four pebbly rock glaciers was estimated by the DC resistivity tomography method (DC resistivity imaging). The electrode configuration followed the Wenner–Schlumberger array, which combines the Wenner and Schlumberger configurations in a profile (e.g. Ishikawa et al., 2001). The sounding profiles were set parallel to the flow line, on which electrodes were placed at 4 or 5 m intervals. The soundings were carried out with the resistivity meter SYSCAL JUNIOR (Iris Instruments, France) from middle July to early August in 2001 and 2002. Two-dimensional modelled DC resistivity distributions were computed with the software RES2DINV ver. 3.4 produced by M. H. Loke.
285
4. Surface features 4.1. Distribution and lithology In the study area, pebbly rock glaciers occur on slopes dominated by shale and/or platy limestone debris (Table 1, Fig. 4b). In contrast, bouldery rock glaciers are located below rockwalls consisting of crystalline or massive sedimentary rocks. Because the areas of shale and platy limestone are restricted, fewer pebbly rock glaciers exist than bouldery rock glaciers (Fig. 4b). Eighty-five percent of the pebbly rock glaciers are located on valley side slopes, reflecting the small size (Table 1, Figs. 2a–c and 5). In particular, the front of three pebbly rock glaciers (A4L, NN2, J2) has not
Fig. 6. Longitudinal profiles of pebbly rock glaciers. A profile of a small but typical bouldery rock glacier is also shown. The vertical scale is elongated to twice the horizontal scale.
286
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Fig. 7. Relationship between the length of pebbly rock glaciers and the maximum height of the source rockwalls. NV=not or sparsely vegetated (probably active) and PV=partly vegetated (probably inactive).
reached the foot of a talus slope (Fig. 2c). Exceptionally long (N 500 m) pebbly rock glaciers (A6, A7, A8) develop in the valley bottom (Fig. 2d). 4.2. Length of a rock glacier and rockwall height The length of rock glaciers relates to the height of the source rockwalls. Seventy-five percent of the pebbly rock glaciers are shorter than 200 m (Table 1, Figs. 5 and 6). Half of such short rock glaciers compose multiple lobes (A1U, A1L, A4U, A4L, A5U, A5L,
BNU, BNL, BN3U, BN3L), but the total length of each lobe is still shorter than 200 m. The source rockwalls of such short rock glaciers are mostly lower than 50 m (Fig. 7). These dimensions contrast with those of active bouldery rock glaciers in the study area, 70–80% of which are longer than 200 m (Hoelzle, 1998; Matsuoka et al., 2005). The source rockwalls are generally higher than 100 m. Pebbly rock glaciers, longer than 200 m (A6, A7, A8, NN8, NN10, NN11, NN12), develop below rockwalls higher than 150 m, where the conglomerate caprock or inclusion of massive limestone can support the high rockwalls (Table 1, Fig. 7). Below a large source rockwall (340 m high), A6 and A7 rock glaciers, the former partly covering the upper part of the latter, constitute a sequence of exceptionally long (c. 1 km) pebbly rock glaciers along a valley floor. 4.3. Thickness and surface gradient of a rock glacier The observed pebbly rock glaciers generally have a low frontal slope compared with the previously known (bouldery) rock glaciers (cf. Burger et al., 1999). The height of frontal slope of the former is smaller than 25 m, except for A4U which lies on a steep slope (Table 1, Fig. 6). Half of the frontal slopes are lower than 10 m. In contrast, active bouldery rock glaciers in the study area mostly have a frontal slope higher than 20 m (Matsuoka et al., 2005). Even the smallest bouldery rock glaciers (PN4 and PN5) have high frontal slopes (c. 20 m: Fig. 6). These results suggest that the pebbly rock glaciers are composed of thinner sediments than
Fig. 8. Downslope variation in the average clast size from the top of a talus slope to a rock glacier. The symbols indicate the clast size on the rock glacier (a) and the talus slope (b). The symbol at 0 m distance shows the average joint spacing of the source rockwall. The lines connect the values on the same survey profile. The solid line (shale) indicates pebbly rock glaciers and the broken line (other lithologies) bouldery rock glaciers. Data for BNU and BNL were obtained along a survey profile.
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
the bouldery rock glaciers, although the frontal height of the former tends to slightly underestimate the thickness, because the upper surface gradually inclines toward the front. The surface gradient of the pebbly rock glaciers ranges from 138 to 298 (average 218), which mostly overlaps with that of bouldery rock glaciers (Matsuoka et al., 2005). The longitudinal profiles of the upper surface of the pebbly rock glaciers, however, are different from those of typical bouldery rock glaciers. The profile of pebbly rock glaciers is convex-up (A1U, A1L, BNL, BW1, NN2, NN10, NN12) or almost flat, whereas that of a typical bouldery rock glacier (PN5) is concave-up (Fig. 6). 4.4. Surface topography The ridge-furrow topography, which is common on bouldery rock glaciers (e.g. Burger et al., 1999), develops only on six pebbly rock glaciers with surface gradient of less than 208 (A6, A7, A8, BN3L, NN1, NN9) (Table 1, Figs. 2 and 5). Despite the absence of the ridge-furrow topography, the upper surface of some pebbly rock glaciers (NN10, NN11, NN12) is undulated because of lobate bulges (Figs. 5b and 6). These bulges may represent a flow unit moving faster than the lower part. Such bulges also develop on talus slopes consisting of shale, for example, those on talus slopes above NN3 and NN4 (Fig. 5b). Many pebbly rock glaciers are situated below talus slopes abundant in fine materials derived from shale or limestone. Traces of debris flows remain on such finerich slopes (Fig. 5). In addition, a veneer of fine debris, originating from a small debris flow or snow avalanche, is often observed on late-lying snow banks covering the upper part of pebbly rock glaciers. In contrast, the latelying snow bank at the top of bouldery rock glaciers commonly shows isolated boulders, although it is occasionally covered by the deposit of a large rockfall (including a dirty snow avalanche).
287
represents a minimum of the original blocks, because the data include fragments shattered in situ after deposition. Thus, pebbly rock glaciers are distinguished from bouldery rock glaciers by the clast size of 15–20 cm in the mean b-axis diameter (Fig. 8a). The maximum b-axis diameter for each lithology is 350 cm (granite), 330 cm (massive limestone), 250 cm (gneiss and schist) and 89 cm (shale). The mean a and c-axes diameters of 25 clasts on pebbly rock glaciers are 8.9–19 cm and 2.5–5.3 cm, respectively, whereas those on bouldery rock glaciers are 34–155 cm and 13–68 cm, respectively. The mean a/c ratios of 25 clasts range from 2.7 to 11 (average 5.7) on pebbly rock glaciers and from 2.0 to 3.4 (average 2.8) on bouldery rock glaciers. The larger ratios for pebbly rock glaciers reflect more tabular clasts of shale. Clasts on the talus slopes above the rock glaciers are mainly cobbles and pebbles regardless of the type of rock glacier. The mean b-axis diameters of 25 clasts are 2.9–8.8 cm on pebbly rock glaciers, and those on bouldery rock glaciers are 6.5–20 cm (granite), 8.8– 18 cm (gneiss and schist) and 2.8–12 cm (massive limestone) (Fig. 8b). The average joint spacings of rockwalls are 9.2–11 cm (granite), 7.4–9.4 cm (gneiss and schist), 4.6–6.5 cm (massive limestone) and 2.0–
4.5. Surface composition Fig. 8 displays the downslope variation in clast size from the top of a talus slope to a rock glacier. The values at 0 m distance show the average joint spacing of the source rockwall. The mean b-axis (intermediateaxis) diameters of 25 clasts are 5.3–13 cm on pebbly rock glaciers mainly consisting of shale, whereas those on bouldery rock glaciers are 24–105 cm for granite, 21–58 cm for gneiss and schist and 24–55 cm for massive limestone. The value for massive limestone
Fig. 9. Stratigraphy of BNU rock glacier. Cobbles and pebbles filled with sandy–silty matrix were entirely frozen below 1 m deep on 3 August, 2000. The shaded area represents frozen debris. The pit did not reach the bottom of permafrost.
288
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
below 2 m depth. The gravimetric ice content of the borecores (c. 10 cm long) was 50% at 4 m depth and 28% at 5 m depth. Ignoring air voids, the volumetric ice content was estimated to be 60% at 4 m depth and 45% at 5 m depth. The borehole did not reach the bedrock at the bottom (5.4 m deep). In addition, five shallow excavations at BNU, BNL and BW revealed the thickness of the uppermost openwork layer varies from 3 cm (one or two fragments) to 50 cm.
3.2 cm (shale), which are comparable with the average clast sizes on the talus slopes (Fig. 8b). In general, the prevailing diameters of surface clasts range from 3 to 25 cm on pebbly rock glaciers (Table 1). Boulders are nearly absent on half of the pebbly rock glaciers. Six pebbly rock glaciers (A3, A4U, A4L, A5U, A5L, J1), derived from limestone, lack boulders and sandy–silty matrix near the surface. Boulders are, however, sparsely distributed on pebbly rock glaciers originating from rockwalls producing blocks and fine clasts (A6, A7, A8, BNL, BN3U, BN3L, BW1, NN3, NN10, NN11, NN12). In particular, BN2 and NN4 are covered with a thin openwork bouldery layer on a matrix-supported layer.
5.2. Seismic refraction soundings The shallow seismic refraction soundings indicated a two-layered structure in three bouldery and nine pebbly rock glaciers (Table 2, see also Fig. 5 for the locations). The first layer with a low P-wave velocity (330–650 m s 1) indicates the active layer and the second layer with a high P-wave velocity (2000–4400 m s 1) permafrost. The computed active layer is 2 to 3 m thick on many active and inactive rock glaciers, but thick active layers (N4 m) are indicated on NN8 and the lower part of BNL (BNL-b), BW2 and NN2. The latter implies degrading permafrost, although these rock glaciers lack a depressed transverse profile indicative of melting ice-supersaturated permafrost (cf. Ikeda and Matsuoka, 2002).
5. Internal structure 5.1. Direct observation The pit and borehole on BNU displayed that platy shale pebbles and cobbles filled with sandy–silty matrix are the major components, except for the uppermost 40 cm of an openwork layer (Fig. 9). The frost table lay at about 1 m depth on August 3, 2000. Below the frost table, the debris was entirely ice-cemented to the bottom of the borehole and lacking visible air voids. A number of small ice lenses (b2 cm thick) were observed
Table 2 Results of seismic refraction and vertical electrical resistivity soundings Site
P-wave stratigraphy First layer V(m s 1)
Second layer
AB
First layer
Second layer
Third layer
Fourth layer
AB/2
V(m s 1)
(m)
U(kgm)
D(m)
U(kgm)
U(kgm)
D(m)
U(kgm)
(m)
4100 3700 4400
55 60 60
19–21 6.8–56
2.4 1.3
260 5.6
4.4 4.2
520 210
17 17
10 1.3
100 100
2.2–2.6 2.1–2.3 4.4–5.8
2800 2900 3400
50 50 40
0.68 4.1–1.4
0.51 1.4
2.2 2.8
5.0 12
3.2 6.7
2.1–2.7 2.0–2.1 2.3–5.3 3.6–4.4 4.8–8.6
3000 2300 2000 2600 3200
42 40 55 45 60
2.2–2.4
2900
50
7.8–29 1.9 0.89 2.6 2.7 2.1–1.6 2.0–0.81 3.8–4.2 4.6–3.2
1.5 1.3 1.4 1.3 0.51 2.7 2.7 4.5 3.8
2.5 0.79 0.81 0.55 2.0 1.2 6.5 8.5 6.4
26 3.7 12 4.1 4.8 8.3 10 16 8.4
1.5 2.0 0.12 0.44 1.8 2.0 5.1 2.0 12
D(m)
Bouldery rock glaciers PN1 550 2.8–2.9 PN5 650 1.3–2.4 PS3* 330 2.6–3.1 Pebbly rock glaciers BNU 390 BNL-a 370 BNL-b 470 BN2** BN3U* 350 BW1 330 BW2 350 NN2 420 NN8* 450 NN10 NN11 NN12 340
DC resistivity stratigraphy
D(m)
20 31
23 22 15 14
23
0.66 0.32
1.1 21 1.8 2.1
0.022
50 100 50 50 80 50 50 50 50 50 100
*Partly vegetated (probably inactive). **A thin openwork bouldery layer on the surface. Symbols: V (P-wave velocity), D (depth of layer base), AB (length of sounding profile) and U (calculated resistivity).
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
5.3. Electrical DC resistivity soundings The vertical soundings showed that the DC resistivities in pebbly rock glaciers were usually one to two orders of magnitude lower than that of typical bouldery rock glaciers at a similar depth (Table 2, Fig. 10). Permafrost in PN5 and PS3 bouldery rock glaciers showed resistivity values of 210–520 kVm. In contrast, the resistivity values of the permafrost in 11 pebbly rock glaciers ranged from 0.44 to 8.5 kVm. In addition, the resistivities in typical bouldery rock glaciers significantly increase below the frost table estimated by the seismic soundings, whereas the values in pebbly rock glaciers only slightly change across the frost table. The sharp decrease in resistivity in BN2 reflects the different clast size between the uppermost bouldery layer and the underlying pebbly layer. The third layer of BNL
289
(6.7 kVm) and NN12 (12 kVm) may also show the resistivities of permafrost. The resistivity imaging displayed a spatial variation in resistivity within a rock glacier. In a longitudinal section, elliptic layers with relatively high resistivity are distinguished from the surroundings by a large resistivity gradient (Figs. 11 and 12). The observed resistivities show a similar range to those obtained from the vertical soundings, although the cores with relatively high resistivity in PN5 (N 1000 kVm) and BW1 (N2 kVm) were not detected by the vertical soundings (Figs. 11 and 12, see also Fig. 10). 5.4. Interpretation The large gradient in resistivity at shallow depth in bouldery rock glaciers certainly indicates the presence
Fig. 10. Apparent resistivity curves and calculated electrical DC resistivities. Parenthetic symbols: b = bouldery rock glacier; p = pebbly rock glacier; and pb = pebbly rock glacier covered with a thin openwork bouldery layer.
290
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Fig. 11. Electrical DC resistivity tomograms of bouldery rock glaciers.
of permafrost (e.g. PN5 in Fig. 10). This contrast is small or even absent in the pebbly rock glaciers, although the P-wave velocities indicate the presence of permafrost. The resistivity values reflect temperature and grain-size distribution (e.g. Hoekstra and McNeill, 1973). For example, the active layer of the bouldery rock glaciers (resistivity 10–50 kVm) has resistivity an order of magnitude higher than that of the pebbly rock glaciers (1–5 kVm), because the interstitial fine materials of the latter hold water that reduces the resistivity. Similarly, the low resistivity of permafrost in the pebbly rock glaciers reflects the higher content of fine debris compared with the bouldery rock glaciers. Permafrost temperature may also affect the resistivity. The observed rock glaciers, lying near the lower limit of mountain permafrost, probably have permafrost close to the melting point (0 to 28C: Arenson et al., 2002). In particular, the decreasing resistivities (BW1, BW2, NN2, NN8) may result from permafrost at the melting point, as suggested by the relatively warm location and/or thick active layer (Table 2, Fig. 10, see also Fig. 5 for the location). The two orders of magnitude difference in resistivity between bouldery and pebbly rock glaciers, however, is mainly attributed to the difference in ice content. The
previous measurements have reported a wide range of resistivity (5–2000 kVm), which probably reflects the variable ice contents from less than 60% in ice-cemented debris (e.g. Fisch et al., 1977; Barsch et al., 1979; Elconin and LaChapelle, 1997; Arenson et al., 2002) to nearly 100% in massive ice (e.g. Vonder Mu¨hll and Holub, 1992; Arenson et al., 2002). Thus, the high resistivity (N 100 kVm) in the bouldery rock glaciers probably results from a highly ice-supersaturated layer (Haeberli et al., 1998), whereas the lower range (b 20 kVm) in the pebbly rock glaciers mostly indicates the lack of massive ice. The two-dimensional images display that permafrost in bouldery rock glaciers have a large spatial variation in resistivity (100–4000 kVm in PN5, 10– 250 kVm in PS3), probably depending on permafrost temperature and ice content (Fig. 11). The high resistivities (N 100 kVm) prevailing under both PN5 and PS3 probably represent highly ice-supersaturated layers. In addition, the very high value (N 1000 kVm) in PN5 indicates even the presence of massive ice (N 10 m thick) at relatively low temperature, whereas the lower range (b30 kVm) at the frontal part of PS3 indicates relatively low ice content and/or high temperature. In comparison, pebbly rock glaciers show much smaller spatial variation in resistivity (Fig. 12). The elliptic layers with relatively high DC resis-
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
291
Fig. 12. Electrical DC resistivity tomograms of pebbly rock glaciers. The depth of the frost table was estimated by seismic soundings.
tivity (4–20 kVm) in pebbly rock glaciers may reflect slightly colder permafrost than the surroundings. The resistivity lower than 2 kVm in the upper part of BNU, the lower part of BNL and the both ends of BW1 probably indicates permafrost at the melting point or even an unfrozen layer. This interpretation is supported by the permafrost temperatures in the borehole on BNU which are nearly 0.08C throughout the year at 3, 4 and 5 m depths (Ikeda et al., 2003).
Also, the low P-wave velocity indicates that the lower part of BNL has a thick unfrozen layer. 6. Processes Fig. 13 compares the morphological characteristics and major geomorphological processes between pebbly and bouldery rock glaciers. The differences between the two depend primarily on lithology of the source rock-
292
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Fig. 13. A schematic diagram of rock control on rock glaciers.
walls. Such a rock control on rock glaciers is discussed in this section. 6.1. Debris input The primary difference between bouldery and pebbly rock glaciers lies in the material composition that inherits characteristics of the source rockwall. The average joint spacing of the rockwalls is much smaller than the dimensions of the clasts on the bouldery rock glaciers (Fig. 8). This indicates that the clasts mainly originate from large rockfalls reflecting the spacing of long and deep cracks. In contrast, the clast size of the pebbly rock glaciers approximates the average joint spacing of the rockwalls (Fig. 8). Thus, the pebbly rock glaciers are mainly fed by small-scale rockfalls. Such pebbly rock glaciers develop mainly below rockwalls closely spaced by joints (cf. Matsuoka and Ikeda, 2001). In particular, shale and phyllite rockwalls are most effective in producing fine materials that constitute the matrix of the active layer (cf. Wahrhaftig and Cox, 1959; Calkin et al., 1987). The variation in clast size from talus slopes to rock glaciers also implies the processes of debris transport. The clast size suddenly increases from the foot of the talus slopes to the head of the bouldery rock glaciers (Fig. 8). This probably results from fall sorting, thereby boulders preferentially accumulating on the foot of talus slopes (e.g. Haeberli et al., 1998). Sorting may be further promoted by sieving and downwashing of fine materials (e.g. Vick, 1981; Haeberli, 1985), eventually resulting in an openwork bouldery layer.
In contrast, the clast size of the pebbly rock glaciers is similar to that on the talus slopes, since cobbles or finer debris preferentially produced from the densely jointed rockwalls are unfavourable for gravitational sorting. Instead, the talus slopes abundant in fine materials are susceptible to debris flow and/or solifluction, which efficiently transport fine materials on the pebbly rock glaciers. Such a transport process is generally lacking on the talus slopes above the active bouldery rock glaciers. The length of rock glaciers is correlated with the size of the source rockwalls (Fig. 7). In general, high, large rockwalls can produce abundant debris feeding long bouldery rock glaciers. In contrast, low rockwalls constrain the downslope growth of pebbly rock glaciers, because they produce a much smaller amount of debris. Shale and platy limestone, however, constitute densely jointed rockwalls that experience more rapid weathering than sparsely jointed rockwalls (e.g. Matsuoka, 1990). Despite the high weathering rate, the small rockwall cannot produce a large amount of debris which supports a large rock glacier (Matsuoka and Ikeda, 2001). In addition, high rockwalls would have intensified debris supply onto bouldery rock glaciers during the past deglaciation (Ballantyne, 2002). 6.2. Permafrost formation A wide range of DC resistivity values, observed in permafrost below bouldery and pebbly rock glaciers, implies a variety of non-glacial ice formation in the rock glaciers. The low resistivity (1–10 kVm) in the pebbly rock glaciers indicates ice-cemented to slightly
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
ice-supersaturated structure (Table 1, Figs. 10 and 12). The direct observation revealed a low resistivity (2 kVm) layer representing frozen debris cemented with ice (Figs. 9 and 12). Such a structure is basically formed by groundwater freezing accompanied by frost heave. Freezing of unconfined groundwater is unlikely to form a highly ice-supersaturated layer or even massive ice (resistivityN 100 kVm) in bouldery rock glaciers (see Fig. 11), which is indicated by the relatively icepoor structure of pebbly rock glaciers. One of the origins for the intensive ice-supersaturation in talusderived rock glaciers is burying snow banks. Longlasting or perennial snow banks, often covered with rockfall deposits, occur at the top of talus-derived rock glaciers (Dyke, 1990; Matsuoka, 1998). Such a debris accumulation can incorporate the snow bank into permafrost, producing an ice-supersaturated layer (e.g. Outcalt and Benedict, 1965; Haeberli et al., 1998; Berthling et al., 2000). The difference in ice content between bouldery and pebbly rock glaciers appears to reflect the different debris accumulation. The large rockwalls above bouldery rock glaciers have potential to produce a thick debris deposit on the snow bank through a large rockfall. The low rockwalls above pebbly rock glaciers are much less favourable for such a large rockfall. Pebbly rock glaciers below a high rockwall may have an ice-supersaturated layer, as indicated by a high resistivity layer (N 10 kVm) in the upper part of NN12 (see Fig. 12). Another origin of the ice-supersaturation is ice injection, which is the major mechanism of pingos and frost blisters (e.g. Mackey, 1973; van Everdingen, 1978; French, 1996). The presence of intrapermafrost water in rock glaciers (e.g. Vonder Mu¨hll, 1992; Vonder Mu¨hll et al., 2003) cannot rule out this possibility, but this process fails to explain the lack of massive ice in pebbly rock glaciers. 6.3. Permafrost deformation Most of the pebbly rock glaciers have a convex-up or flat longitudinal profile, thereby the surface gradient increases or maintains towards the front (Fig. 6). The increasing surface gradient results in the predominance of extending flow, which is responsible for the lack of transverse ridge–furrow topography produced by compressive flow (Haeberli et al., 1979). In contrast, many bouldery rock glaciers decline towards the front, where the ridge–furrow topography is common. The difference in the surface topography between bouldery and pebbly rock glaciers depends mainly on the length and location. Bouldery rock glaciers often develop into the
293
valley bottom where the surface gradient declines, whereas pebbly rock glaciers usually terminate within (or at the foot of) a valley-side slope. The exceptionally long pebbly rock glaciers developing along the valley bottom (A6, A7, A8) have the ridge-furrow topography (Fig. 2d). The longitudinal compression causes thickening of rock glaciers. The thickened ice-supersaturated layer results in increasing volumetric ice content. Thus, the ice-supersaturation is probably promoted by long-lasting compression, which is common in long bouldery rock glaciers. Such thickening explains also why many bouldery rock glaciers have higher frontal slopes (N 20 m) than pebbly rock glaciers. 6.4. Permafrost preservation Near the lower limit of mountain permafrost, the active layer composition also affects the preservation of permafrost, which controls the altitudes of the terminus of rock glaciers. The openwork structure of the active layer of bouldery rock glaciers permits intensive cooling of the underlying permafrost and the storage of cold air in winter (e.g. Keller and Gubler, 1993; Hoelzle et al., 1999; Matsuoka and Ikeda, 2003). Below a thick snow cover the surface of active pebbly rock glaciers maintain higher temperatures than that of active bouldery rock glaciers (Ikeda et al., 2003). Thus, bouldery rock glaciers have a potential to extend to lower altitudes. The frontal altitude of two active bouldery rock glaciers (PN1, PN2), 2540 m ASL, is lower than that of a nearby relict pebbly rock glacier (BN3L), 2610 m ASL (Fig. 5a). 6.5. Active layer modification The mobility of the active layer of rock glaciers depends on the presence of fine materials. Matrix-free blocks on bouldery rock glaciers move passively by creep of the underlying permafrost. The sharp frontal edge is maintained on active bouldery rock glaciers, because individual boulders in the active layer only fall or slide on the frontal slope. In contrast, rounded frontal edge of pebbly rock glaciers indicates solifluction or other soil transport in the matrix-supported layer. In addition, solifluction lobes develop from the side of a pebbly rock glacier in the Furggenta¨lti, Swiss Alps (Krummenacher and Budmigner, 1992), and patterned ground occurs on some pebbly rock glaciers in the Brooks Range, Alaska (Calkin et al., 1987). Such a soil movement may gradually decrease the thickness of pebbly rock glaciers.
294
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
7. Conclusions
Acknowledgements
The difference in the source rockwall geology, which controls the clast size, determines the two types of rock glaciers. Bouldery rock glaciers have the active layer composed of matrix-free boulders originating from resistant rocks (e.g. crystalline rocks and massive limestone), whereas pebbly rock glaciers consist mainly of pebbles and cobbles filled with sandy–silty matrix from less resistant rocks (e.g. shale and platy limestone). Pebbly rock glaciers are distinguished from bouldery rock glaciers by the clast size of 15–20 cm in the mean b-axis diameter. The clast size controls transport processes from the source to the front of a rock glacier, resulting in different shape, size and internal structure between pebbly and bouldery rock glaciers. The matrix-free boulders (20–100 cm in the mean baxis diameter) imply that large rockfalls accompanied by fall sorting of debris are the major process of the debris input on bouldery rock glaciers. In contrast, the pebbles and cobbles (5–10 cm in the mean b-axis diameter) filled with sandy–silty matrix imply that small-scale rockfalls, aided by debris flow and/or solifluction, feed the pebbly rock glaciers. The fine materials keep the active layer of pebbly rock glaciers susceptible to solifluction, which results in the rounded frontal edge. Seventy-five percent of the pebbly rock glaciers are shorter than 200 m, whereas bouldery rock glaciers are typically longer than 200 m. The source rockwalls of such short pebbly rock glaciers are lower than 50 m, whereas those of bouldery rock glaciers are generally higher than 100 m. The small exposure of the less resistant source rockwall strongly constrains debris supply. As a result, pebbly rock glaciers usually terminate within a valley-side slope, whereas many bouldery rock glaciers extend into the valley bottom. The location of pebbly rock glaciers results in the lack of compressive flow, which is responsible for the low frontal slope (V 20 m) and the lack of transverse ridge–furrow topography. The pit–borehole stratigraphy and low DC resistivities (b10 kVm) indicate ice-cemented or slightly supersaturated permafrost in the pebbly rock glaciers. Such a structure may largely originate from groundwater freezing. In contrast, highly ice-supersaturated structure (or massive ice) indicated by high DC resistivities (N100 kVm) in bouldery rock glaciers appears to originate from buried snow banks, which have been buried with large rockfall deposits from the large source rockwall.
This study was financially supported by the Grantin-Aid (No. 15500676) for Scientific Research and Grant-in-Aid for JSPS Fellows from the Ministry of Education, Science and Culture, Japan and by the Sasakawa Scientific Research Grant from the Japan Scientific Society. We acknowledge F. Keller for logistical help, K. Hirakawa, T. Watanabe, M. Aoyama, K. Fukui, Y. Sawada and S. Iwasaki for their cooperation in the field, and Mi. Abe, Ma. Abe, T. Date and Y. Eda for field assistance. References Arenson, L., Hoelzle, M., Springman, S., 2002. Borehole deformation measurements and internal structures of some rock glaciers in Switzerland. Permafr. Periglac. Process. 13, 117 – 135. Ballantyne, C.K., 2002. Paraglacial geomorphology. Quat. Sci. Rev. 21, 1935 – 2017. Barsch, D., 1977. Nature and importance of mass-wasting by rock glaciers in alpine permafrost environments. Earth Surf. Processes 2, 231 – 245. Barsch, D., Fierz, H., Haeberli, W., 1979. Shallow core drilling and bore-hole measurements in the permafrost of an active rock glacier near the Grubengletscher Wallis, Swiss Alps. Arct. Alp. Res. 11, 215 – 228. Berthling, I., Etzelmu¨ller, B., Isaksen, K., Sollid, J.L., 2000. Rock glaciers on Prins Karls Forland: II. GPR soundings and the development of internal structures. Permafr. Periglac. Process. 11, 357 – 369. Burger, K.C., Degenhardt Jr., J.J., Giardino, J.R., 1999. Engineering geomorphology of rock glaciers. Geomorphology 31, 93 – 132. Calkin, P.E., Haworth, L.A., Ellis, J.M., 1987. Rock Glaciers of Central Brooks Range, Alaska, U.S.A.. In: Giardino, J.R., Shroder Jr., J.F., Vitek, J.D. (Eds.), Rock Glaciers. Allen and Unwin, London, pp. 65 – 82. Chueca, J., 1992. A statistical analysis of the spatial distribution of rock glaciers, Spanish central pyrenees. Permafr. Periglac. Process. 3, 261 – 265. Dyke, A.S., 1990. A lichenometric study of Holocene rock glaciers and neoglacial moraines, Frances Lake Map Area, southeastern Yukon Territory and Northwest Territories. Pap. - Geol. Surv. Can. 394, 1 – 33. Elconin, R.F., LaChapelle, E.R., 1997. Flow and internal structure of a rock glacier. J. Glaciol. 43, 238 – 244. Evin, M., 1987. Lithology and fracturing control of rock glaciers in southwestern Alps of France and Italy. In: Giardino, J.R., Shroder Jr., J.F., Vitek, J.D. (Eds.), Rock Glaciers. Allen and Unwin, London, pp. 83 – 106. Fisch Sr., W., Fisch Jr., W., Haeberli, W., 1977. Electrical D.C. resistivity soundings with long profiles on rock glaciers and moraines in the Alps of Switzerland. Z. Gletsch.kd. Glazialgeol. 13, 239 – 260. Frauenfelder, R., Ka¨a¨b, A., 2000. Towards a paleoclimatic model of rock glacier formation in the Swiss Alps. Ann. Glaciol. 31, 281 – 286. Frauenfelder, R., Haeberli, W., Hoelzle, M., Maisch, M., 2001. Using relict rockglaciers in GIS-based modelling to reconstruct Younger
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296 Dryas permafrost distribution patterns in the Err–Julier area Swiss Alps. Nor. Geogr. Tidsskr. 55, 195 – 202. Frauenfelder, R., Haeberli, W., Hoelzle, M., 2003. Rockglacier occurrence and related terrain parameters in a study area of the Eastern Swiss Alps. Proc. 8th Int. Conf. Permafrost. Balkema, Lisse, Netherlands, pp. 253 – 258. French, H.M., 1996. The Periglacial Environment, 2nd ed. Longman, London. Frich, P., Brandt, E., 1985. Holocene talus accumulation rates, and their influence on rock glacier growth. A case study from Igpik, Disko, West Greenland. Geogr. Tidsskr. 85, 32 – 44. Giradino, J.R., Vitek, J.D., 1988. The significance of rock glaciers in the glacial–periglacial landscape continuum. J. Quat. Sci. 3, 97 – 103. Haeberli, W., 1985. Creep of mountain permafrost: internal structure and flow of alpine rock glaciers. Mitt. Versuchsanst. Wasserb., Hydrol. Glaziol. ETH Zu¨r. 77 (142 pp.). Haeberli, W., Patzelt, G., 1982. Permafrostkartierung im Gebiet der ¨ tztaler Alpen. Z. Hochebenkar–Blockgletscher, Obergurgl, O Gletsch.kd. Glazialgeol. 18, 127 – 150. Haeberli, W., Vonder Mu¨hll, D., 1996. On the characteristics and possible origins of ice in rock glacier permafrost. Z. Geomorphol. N. F. Suppl.Bd. 104, 43 – 57. Haeberli, W., King, L., Flotron, A., 1979. Surface movement and lichen-cover studies at the active rock glacier near the Grubengletscher, Wallis, Swiss Alps. Arct. Alp. Res. 11, 421 – 441. Haeberli, W., Evin, M., Tenthorey, G., Kensen, H.R., Hoelzle, M., Keller, F., Vonder Mu¨hll, D., Wagner, S., Pelfini, M., Smiraglia, C., 1992. Permafrost research sites in the Alps: excursions of the international workshop on permafrost and periglacial environments in mountain areas. Permafr. Periglac. Process. 3, 189 – 202. Haeberli, W., Hoelzle, M., Keller, F., Vonder Mu¨hll, D., Wagner, S., 1998. Ten years after the drilling through the permafrost of the active rock glacier Murte`l, eastern Swiss Alps: answered questions and new perspectives. Proc. 7th Int. Conf. Permafrost. Centre d’e´tudes Nordiques, Sainte-Foy, Canada, pp. 403 – 410. Haeberli, W., Ka¨a¨b, A., Wagner, S., Vonder Mu¨hll, D., Geissler, P., Haas, J.N., Glatzel-Mattheier, H., Wagenbach, D., 1999. Pollen analysis and 14C age of moss remains in a permafrost core recovered from the active rock glacier Murte`l–Corvatsch, Swiss Alps: geomorphological and glaciological implications. J. Glaciol. 45, 1 – 8. Hoekstra, P., McNeill, D., 1973. Electromagnetic probing of permafrost. Proc. 2nd Int. Conf. Permafrost. National Academy of Sciences, Washington, pp. 517 – 526. Hoelzle, M., 1998. Rock glaciers, Upper Engadin, Switzerland. IPA Data and Information Working Group, Comp. Circumpolar Active-layer Permafrost System (CAPS), ver. 1.0. National Snow and Ice Data Center, Boulder, CD-ROM, ROM. Hoelzle, M., Wegmann, M., Krummenacher, B., 1999. Miniature temperature dataloggers for mapping and monitoring of permafrost in high mountain areas: first experience from the Swiss Alps. Permafr. Periglac. Process. 10, 113 – 124. Humlum, O., 2000. The geomorphic significance of rock glaciers: estimates of rock glacier debris volumes and headwall recession rates in West Greenland. Geomorphology 35, 41 – 67. Hunter, J.A.M., 1973. The application of shallow seismic methods to mapping of frozen surficial materials. Proc. 2nd Int. Conf. Permafrost. National Academy of Sciences, Washington, pp. 527 – 535.
295
Ikeda, A., Matsuoka, N., 2002. Degradation of talus-derived rock glaciers in the Upper Engadin, Swiss Alps. Permafr. Periglac. Process. 13, 145 – 161. Ikeda, A., Matsuoka, N., Ka¨a¨b, A., 2003. A rapidly moving small rock glacier at the lower limit of the mountain permafrost belt in the Swiss Alps. Proc. 8th Int. Conf. Permafrost. Balkema, Lisse, Netherlands, pp. 455 – 460. Ishikawa, M., Watanabe, T., Nakamura, N., 2001. Genetic differences of rock glaciers and the discontinuous mountain permafrost zone in Kanchanjunga Himal, eastern Nepal. Permafr. Periglac. Process. 12, 243 – 253. Ka¨a¨b, A., Gudmundsson, G.H., Hoelzle, M., 1998. Surface deformation of creeping mountain permafrost. Photogrammetric investigations on Murte`l rock glacier, Swiss Alps. Proc. 7th Int. Conf. Permafrost. Centre dTe´tudes Nordiques, Sainte-Foy, Canada, pp. 531 – 537. Keller, F., Gubler, H., 1993. Interaction between snow cover and high mountain permafrost Murte`l/Corvatsch, Swiss Alps. Proc. 6th Int. Conf. Permafrost, vol. 1. South China University Technology Press, Wushan, China, pp. 332 – 337. Krummenacher, B., Budmigner, K., 1992. Monitoring of periglacial phenomena in the Furggenta¨lti Swiss Alps. Permafr. Periglac. Process. 3, 149 – 155. Mackey, J.R., 1973. The growth of pingos, western Arctic coast, Canada. Can. J. Earth Sci. 10, 979 – 1004. Matsuoka, N., 1990. The rate of bedrock weathering by frost action: field measurements and a predictive model. Earth Surf. Process. Landf. 15, 73 – 90. Matsuoka, N., 1998. Monitoring rockwall instability in the Murtel–Corvatsch region, Upper Engadin. Beitra¨ge aus der Gebirgs-Geomorphologie, Mitteilungen der Versuchsanstalt fu¨r Wasserbau, Hydrologie und Glaziologie ETH Zu¨rich, vol. 158, pp. 7 – 14. Matsuoka, N., Sakai, H., 1999. Rockfall activity from an alpine cliff during thawing periods. Geomorphology 28, 309 – 328. Matsuoka, N., Ikeda, A., 2001. Geological control on the distribution and characteristics of talus-derived rock glaciers. Annu. Rep. Inst. Geosci., Univ. Tsukuba 27, 11 – 16. Matsuoka, N., Ikeda, A., 2003. Intensive cooling through an open-work blocky layer: ground temperature monitoring in the Swiss Alps. Annu. Rep. Inst. Geosci., Univ. Tsukuba 29, 21 – 25. Matsuoka, N., Hirakawa, K., Watanabe, T., Moriwaki, K., 1997. Monitoring of periglacial slope processes in the Swiss Alps: the first two years of frost shattering, heave and creep. Permafr. Periglac. Process. 8, 155 – 177. Matsuoka, N., Ikeda, A., Date, T., 2005. Morphometric analysis of solifluction lobes and rock glaciers in the Swiss Alps. Permafr. Periglac. Process. 16, 99 – 113. Outcalt, S.I., Benedict, J.B., 1965. Photo-interpretation of two types of rock glaciers in the Colorado Front Range, U.S.A. J. Glaciol. 5, 849 – 856. Palmer, D., 1986. Refraction Seismics. Geophysical Press, London. van Everdingen, R.O., 1978. Frost mounds on Bear Rock, near Fort Norman, N.W.T. 1975-1976. Can. J. Earth Sci. 15, 263 – 276. Vick, S.G., 1981. Morphology and the role of landsliding in formation of some rock glaciers in the Mosquito Range, Colorado. Geol. Soc. Amer. Bull. 92, 75 – 84. Vonder Mu¨hll, D.S., 1992. Evidence of intrapermafrost groundwater flow beneath an active rock glacier in the Swiss Alps. Permafr. Periglac. Process. 3, 169 – 173.
296
A. Ikeda, N. Matsuoka / Geomorphology 73 (2006) 279–296
Vonder Mu¨hll, D.S., Holub, P., 1992. Borehole logging in alpine permafrost, Upper Engadin, Swiss Alps. Permafr. Periglac. Process. 3, 125 – 132. Vonder Mu¨hll, D.S., Hauck, C., Lehmann, F., 2000. Verification of geophysical models in Alpine permafrost by borehole information. Ann. Glaciol. 31, 300 – 306.
Vonder Mu¨hll, D.S., Arenson, L.U., Springman, S.M., 2003. Temperature conditions in two alpine rock glaciers. Proc. 8th Int. Conf. Permafrost. Balkema, Lisse, Netherlands, pp. 1195 – 2000. Wahrhaftig, C., Cox, A., 1959. Rock glaciers in the Alaska Range. Bull. Geol. Soc. Am. 70, 383 – 436.